ClimateLevel 6 of 6 territoryTerritory, level 6 of 6Calculated with 13 entries, levels 1 and 4Database structure 
5Organisations
4Occupations and activities
3Uses
2Machine operations
1Calculation factorsPublic servants in France, work, 1 year
The 5.85 million staff of the three branches of the French civil service, teachers, healthcare workers, administrative and technical staff, for the share that uses AI at work. Projet Celsius estimate.
- "Office employee, 1 year" accounts for 29% of emissions in this scenario.
- Light use: 84 tCO2e; heavy use: 3.3 ktCO2e, with the other settings unchanged.
- Out of scope: the AI services of public administrations, personal use by staff, server manufacturing and model training.
Sentence to citePublic servants in France, work, 1 year, Projet Celsius estimate: 640 tCO2e per population and per year, for 1.4 GWh of electricity, servers worldwide (458 gCO2e/kWh).
Breakdown of emissions by item
Occupations and activities
- Office worker, 1 year29 %
- Consultant, 1 year15 %
- Developer, 1 year14 %
- Field-based occupation, 1 year13 %
- Teacher, 1 year7 %
- 7 others22 %
Uses
- Web search with an AI21 %
- Quick question9 %
- Write meeting minutes8 %
- Code suggestion8 %
- Presentation of 10 slides6 %
- 19 others48 %
Machine operations and data centre
- Read 1,000 tokens42 %
- Write, with reasoning27 %
- Write 1,000 tokens10 %
- PUE of the data centre9 %
- Generate an image5 %
- 3 others7 %
640 tCO2e in total: 1.4 GWh of electricity × 458 gCO2e/kWh (servers worldwide). Each bar splits the emissions between the entries of a category.
Inputs of the calculation
PUE of the data centre, Read 1,000 tokens, Write, with reasoning, Write 1,000 tokens, Re-read from cache, Generate an image, Second of video, Quick question, Proofread a page, Write meeting minutes, Presentation of 10 slides, Web search (benchmark), Query a PDF, Web search with an AI, Code suggestion, Consultant, 1 year, Researcher, 1 year, Developer, 1 year, Office worker, 1 year, Executive assistant, 1 year, Teacher, 1 year, Field-based occupation, 1 year
How the calculation is built
Insee headcount at the end of 2024; the split by occupation is an assumption.
Comparison with other territories
Characteristics
| Unit | 1 population, 1 year |
|---|---|
| Default location | world, 458 gCO2e/kWh (Ember 2025, life cycle); 11 other locations to choose from |
| Counted | the use of AI at work by public servants, occupation by occupation, data centre included |
| Not counted | the AI services of public administrations, personal use by staff, server manufacturing and model training |
| Status | estimated |
All the characteristics
| Identifier | IA-0515, projetcelsius.com/en/ai/database/IA-0515/ |
|---|---|
| Version | 0.1 of 2 November 2026 |
| Category | Territories, cities, territories, global services |
| Choice of the selected value | database calculation from the entries this one is made of, at the quantities of the unit |
| Sources | calculation of the database |
| Licence | CC BY 4.0 |
Usage parameters
Advanced settings
Result for the chosen case
220 days, 5.4 seconds per video, servers worldwide: 640 tCO2e, 1.4 GWh.
Breakdown by item
Quantity, unit value, contribution and nature of the data for each item
| Item | Quantity | Unit value | Contribution | Share | Nature of the data |
|---|---|---|---|---|---|
| Office worker, 1 year | × 319,000 | 1.3 kWh | 410 MWh | 29 % | assumption |
| Consultant, 1 year | × 100,000 | 2.1 kWh | 210 MWh | 15 % | assumption |
| Developer, 1 year | × 43,800 | 4.3 kWh | 190 MWh | 14 % | assumption |
| Field-based occupation, 1 year | × 211,000 | 480 Wh | 100 MWh | 7 % | assumption |
| Teacher, 1 year | × 163,000 | 570 Wh | 92 MWh | 7 % | assumption |
| Executive assistant, 1 year | × 56,300 | 1.4 kWh | 78 MWh | 6 % | assumption |
| Researcher, 1 year | × 50,000 | 1.5 kWh | 74 MWh | 5 % | assumption |
| Field-based occupation, 1 year | × 107,000 | 480 Wh | 51 MWh | 4 % | assumption |
| Manager or executive, 1 year | × 55,000 | 810 Wh | 45 MWh | 3 % | assumption |
| Accountant, 1 year | × 40,000 | 830 Wh | 33 MWh | 2 % | assumption |
| Human resources, 1 year | × 40,000 | 720 Wh | 29 MWh | 2 % | assumption |
| Lawyer, 1 year | × 25,000 | 1.1 kWh | 28 MWh | 2 % | assumption |
| Field-based occupation, 1 year | × 58,100 | 480 Wh | 28 MWh | 2 % | assumption |
| Communications officer, 1 year | × 15,000 | 1.7 kWh | 26 MWh | 2 % | assumption |
Entries that use it
No other entry in the database is calculated from this one.
Sources of the calculation
31 publications, for this entry and those used to calculate it, with the chosen parameters.
The result is calculated by Projet Celsius. The sources below support its uses, its factors and its conversion; they do not all measure this total.
Uses and assumptions
These publications shed light on AI tasks and uses. They do not measure the carbon footprint of the case calculated.
Anthropic · June 2025
Role and limits : Anthropic
What the publication covers. an agent uses about 4 times the tokens of a conversation, a multi-agent system about 15 times
What it does not cover. electricity
Reading by the database. Publisher orders of magnitude, with no absolute value
Deep research → : Re-read from cache →
The context re-read from cache at each of the 40 steps, 40,000 tokens on average
Scope of the figure used. server, convention C4
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Anthropic · February 2025
Role and limits : Anthropic
Author. Kunal Handa et al. (Anthropic)
What the publication covers. Claude.ai conversations matched to O*NET tasks and occupations
What it does not cover. API use and enterprise plans
Reading by the database. A single assistant, more oriented towards code than average
Published observation on uses
Scope of the figure used. Share of Claude.ai conversations linked to computing and mathematics occupations
Limit. Support for the place of code among users of a code-oriented assistant
Published observation on uses
Scope of the figure used. Share of conversations related to education and libraries
Limit. Figure read from the illustration on the publisher's page
Anthropic · March 2026
Anthropic Economic Index report: Learning curves ↗ (new tab)
Role and limits : Anthropic
What the publication covers. Claude.ai conversations from 5 to 12 February 2026 and API use
What it does not cover. the other assistants
Reading by the database. Tracks how the uses of a single assistant evolve
Published observation on uses
Scope of the figure used. Share of Claude.ai conversations linked to computing and mathematics, February 2026
Limit. Code remains the first use, a little less concentrated than in 2025
Apec · June 2025
Role and limits : Apec
What the publication covers. frequency of generative AI use at work, 2,000 private-sector managers surveyed in March 2025
What it does not cover. non-managers and the number of requests
Reading by the database. Recent French survey on frequency of use at work; copy read on a specialist press site
Manager or executive, 1 year →
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Communications officer, 1 year →
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Published observation on uses
Scope of the figure used. Private-sector managers who use generative AI at work every day or almost, March 2025
Limit. Support for the typical level: the median manager user was weekly in early 2025
Published observation on uses
Scope of the figure used. Managers who use it at least once a week without being daily, March 2025
Limit. Support for the typical level: the median manager user was weekly in early 2025
Published observation on uses
Scope of the figure used. Managers who use it at least once a month without being weekly, March 2025
Limit. Support for the typical level: the median manager user was weekly in early 2025
Published observation on uses
Scope of the figure used. Managers who never use it, March 2025 (13% less than once a month)
Limit. Support for the typical level: the median manager user was weekly in early 2025
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Support for the "Typical use" setting
Scope of the figure used. The median user among those who use AI, end of 2026: they turn to it every working day for part of their questions, writing and reading, about 5 to 15 requests per day in an office job. In an organisation or a territory, it also sets the share of employees who use it (25%).
Limit. This setting remains an assumption of the database, not a measurement of the occupation.
Crédoc · 2026
Role and limits : Crédoc
Author. Crédoc for Arcep, Arcom, the CGE and the ANCT
What the publication covers. use of generative AI by people aged 12 and over in France in 2025: who, for what, how often, paid or not
What it does not cover. the number of requests
Reading by the database. The reference survey in France, 4,145 people
Manager or executive, 1 year → : Quick question →
About 5 questions per day, assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Quick question →
About 5 questions per day, assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Quick question →
About 5 questions per day, assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Quick question →
About 5 questions per day, assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Quick question →
About 5 questions per day (employment law, procedures), assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Quick question →
About 5 questions per day (accounting rules, tax), assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Users who turn first to generative AI to look for information (55% through a search engine)
Limit. Survey of 2025 published in 2026, 4,145 people; support for the frequencies and the use
Published observation on uses
Scope of the figure used. Working people who use generative AI in a professional setting (42% of French people for personal use)
Limit. Survey of 2025 published in 2026, 4,145 people; support for the frequencies and the use
Office employee, 1 year → : Quick question →
About 8 information needs per day; share done with AI based on the share of users who turn to an AI first to look something up (Crédoc, Baromètre du numérique, 2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Quick question →
About 6 questions per day, assumption; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Quick question →
About 660 questions and 360 report comments per year, assumption; share done with AI based on the Crédoc Baromètre du numérique (edition 2026, data 2025), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
The 17 other publications in this section
DEPP · November 2025
Role and limits
Author. DEPP, ministère de l'Éducation nationale
What the publication covers. working time reported over a typical week: teaching, lesson preparation, marking, other activities
What it does not cover. the use of digital tools and AI
Reading by the database. French public reference source on teachers' time
Published observation on uses
Scope of the figure used. Hours of classroom teaching reported by secondary school teachers, for 18 hours of service for certified teachers
Limit. Supports the frequency of lesson preparation
Published observation on uses
Scope of the figure used. Share of time spent preparing lessons among certified teachers, for about 42 hours of work a week
Limit. Supports the frequency of lesson preparation
EnronSent
Role and limits
Author. Will Styler
What the publication covers. 96,106 messages sent by Enron employees, 13,810,266 words
What it does not cover. the messages received
Reading by the database. Professional emails from 2000 and 2001: an order of magnitude, not a norm
Write an email → : Write, with reasoning →
The email written, 150 words
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Write an email → : Read 1,000 tokens →
Length convention: an email
Scope of the figure used. A typical email: the sent messages of the Enron corpus average about 144 words; figure read in an excerpt, notice not verified
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Write an email → : Write, with reasoning →
Length convention: an email
Scope of the figure used. A typical email: the sent messages of the Enron corpus average about 144 words; figure read in an excerpt, notice not verified
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Epoch AI · February 2025
Role and limits
What the publication covers. the calculation from the hardware and the assumed size of the model
What it does not cover. actual measurements
Reading by the database. A public, transparent calculation
Reference of the selected value
Scope of the figure used. GPT-4o, 500 tokens written, full scope
Limit. A public calculation, from the hardware and the assumed size of the model
Conversation of 10 exchanges → : Re-read from cache →
The context re-read from cache at each turn: the previous exchanges, and the document already in the conversation if there is one
Scope of the figure used. server, convention C4
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Eurostat · June 2026
Individuals - use of generative AI tools (isoc_ai_iaiu) ↗ (new tab)
Role and limits
Author. Eurostat, survey on ICT use in households
What the publication covers. share of people aged 16 to 74 who used generative AI in the last 3 months, at work, privately, for studies, by country and category, 2025 data
What it does not cover. the frequency of use
Reading by the database. Harmonised European survey, first year
Published observation on uses
Scope of the figure used. France, IT professionals, used generative AI at work in the last 3 months, 2025
Limit. Share of users, which depends on the organisation and the population; here, support for the gap between occupations
Published observation on uses
Scope of the figure used. France, 16 to 74 years, used generative AI at work in the last 3 months, 2025 (European Union: 15.4%)
Limit. Share of users, which depends on the organisation and the population; here, support for the gap between occupations
Published observation on uses
Scope of the figure used. France, employed and unemployed people, used generative AI at work in the last 3 months, 2025
Limit. Share of users, which depends on the organisation and the population; here, support for the gap between occupations
Gallup · June 2025
Teaching for Tomorrow: Unlocking Six Weeks a Year With AI ↗ (new tab)
Role and limits
Author. Gallup and Walton Family Foundation
What the publication covers. use of AI by American primary and secondary school teachers, by task and frequency, year 2024-2025
What it does not cover. teachers outside the United States
Reading by the database. American survey; France would be lower (OECD TALIS survey, not reviewed)
Published observation on uses
Scope of the figure used. American primary and secondary school teachers who used an AI tool for their work in 2024-2025
Limit. Supports the teacher's use of AI; American survey, France is lower
Published observation on uses
Scope of the figure used. Teachers who prepare their lessons with AI at least once a month (about 17% every week, 3% every day)
Limit. Supports the teacher's use of AI; American survey, France is lower
Published observation on uses
Scope of the figure used. Teachers who make their exercise sheets with AI at least once a month
Limit. Supports the teacher's use of AI; American survey, France is lower
Published observation on uses
Scope of the figure used. Teachers who mark or comment on pupils' papers with AI at least once a month
Limit. Supports the teacher's use of AI; American survey, France is lower
Published observation on uses
Scope of the figure used. Time saved reported by teachers who use AI every week
Limit. Self-reported; not used in the calculation
Teacher, 1 year → : Prepare a lesson →
About 8 lessons prepared or redone per week, for 18 hours of teaching and a quarter of working time spent on preparation (DEPP, 2025); share done with AI based on the 37% of teachers who prepare their lessons with AI at least once a month (Gallup, 2025).
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Write a page →
About 4 exercise sheets or assessments per week, assumption; share done with AI based on the 33% who make their sheets with AI at least once a month (Gallup, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Proofread a page →
About 60 marked papers per week, marking being weekly or more often in O*NET; very low share done with AI, only 16% marking with AI at least once a month (Gallup, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : 10-slide presentation →
About 1 slide deck per week, assumption; share done with AI based on Gallup (2025).
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Role and limits
What the publication covers. about 6,000 characters of code read for a suggestion
What it does not cover. electricity
Reading by the database. A post of 2023, updated in 2024: the current size is not published
Code suggestion → : Read 1,000 tokens →
The code around the cursor, about 6,000 characters or 125 lines, read by a small model
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Google · 2026
Role and limits
What the publication covers. up to hundreds of sites browsed and pages processed per search
What it does not cover. electricity
Reading by the database. A presentation page: a ceiling, not an average
Deep research → : Read 1,000 tokens →
The web pages read, about a hundred by default: Google announces up to hundreds of sites
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Insee · February 2026
L'emploi dans la fonction publique en 2024 (Insee Première n° 2094) ↗ (new tab)
Role and limits
Author. Insee, Esbelin, Moreau and Touzé
What the publication covers. staff of the three branches of the civil service at the end of 2024, by branch, category and status
What it does not cover. subsidised contracts, included in the total used
Reading by the database. The reference for public employment
Public servants → : Manager or executive, 1 year →
Executives and managers in the civil service: 110,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Consultant, 1 year →
Engineers, consultants and research officers in the civil service: 200,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Legal adviser, 1 year →
Lawyers in the civil service: 50,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Researcher, 1 year →
Researchers in the civil service: 100,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Communications officer, 1 year →
Communication and marketing officers in the civil service: 30,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Human resources, 1 year →
Human resources officers in the civil service: 80,000; AI use 2 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Accountant, 1 year →
Accountants and finance professionals in the civil service: 100,000; AI use 1.6 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Developer, 1 year →
Developers in the civil service: 70,000; AI use 2.5 times that of the average worker
Scope of the figure used. agent hours per day
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Office employee, 1 year →
Office employees and technicians in the civil service: 850,000; AI use 1.5 times that of the average worker
Scope of the figure used. queries and responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Executive assistant, 1 year →
Assistants and secretaries in the civil service: 150,000; AI use 1.5 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Teacher, 1 year →
Teachers in the civil service: 1.00 million; AI use 0.65 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Field occupation, 1 year →
Skilled field occupations (care workers, technicians, social work) in the civil service: 1.30 million; AI use 0.65 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Field occupation, 1 year →
Service and sales occupations in the civil service: 950,000; AI use 0.45 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Public servants → : Field occupation, 1 year →
Manual workers, farmers and manual occupations in the civil service: 860,900; AI use 0.27 times that of the average worker
Scope of the figure used. tasks handed to the AI, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Microsoft Research · July 2025
Working with AI: Measuring the Applicability of Generative AI to Occupations ↗ (new tab)
Role and limits
Author. Kiran Tomlinson, Sonia Jaffe, Will Wang, Scott Counts, Siddharth Suri (Microsoft)
What the publication covers. 200,000 conversations with Bing Copilot matched to O*NET activities; applicability score by occupation, in open data (version 7 of October 2026)
What it does not cover. actual adoption by occupation
Reading by the database. Measures what AI is used for and succeeds at, not who uses it
Manager or executive, 1 year →
Published observation on uses
Scope of the figure used. Generative AI applicability score for executives (code 11-1011), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Manager or executive, 1 year →
Published observation on uses
Scope of the figure used. Generative AI applicability score for general and operations managers (code 11-1021), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Manager or executive, 1 year → : Write an email →
About 40 emails per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Proofread a page →
About 2 texts to rework per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Conversation of 10 exchanges →
About 3 decisions or interviews to prepare per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Write meeting minutes →
About 3 sets of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Write a page →
About 2 pages per week (note, strategy), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Analyse a spreadsheet →
About 2 dashboards to analyse per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : 10-slide presentation →
About 1 presentation per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Manager or executive, 1 year → : Deep research →
About 1 in-depth question per month (market, competitor), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, context re-read, report, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for management analysts (code 13-1111), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Consultant, 1 year → : 10-slide presentation →
About 2 presentations per week, the typical deliverable of the occupation, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Write a page →
About 2 pages per day (note, proposal, deliverable), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Write an email →
About 20 emails per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Query a PDF →
About 3 long documents to query per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Write meeting minutes →
About 4 sets of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Literature search →
About 1 literature search per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Deep research →
About 2 in-depth questions per month, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, context re-read, report, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Conversation of 10 exchanges →
About 3 reflection sessions per week (structuring a problem), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for lawyers (code 23-1011), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Legal adviser, 1 year → : Write a page →
About 1 page per day (note, opinion, clause), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Proofread a page →
About 3 texts to rework per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Write an email →
About 20 emails per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Summarise 30 pages →
About 3 decisions or reports to summarise per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Query a PDF →
About 3 long case files to query per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Literature search →
About 1 literature search per week (case law, legal scholarship), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Translate a page →
About 2 pages to translate per week (contracts in English), assumption; share done with AI set high, translation being among the tasks most often delegated to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Write meeting minutes →
About 2 sets of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Query a PDF →
About 5 scientific articles queried per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Summarise 30 pages →
About 3 articles or reports to summarise per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Literature search →
About 1 bibliographic search per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Deep research →
About 2 quick literature reviews per month, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, context re-read, report, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Write a page →
About 1 page per day (article, project, report), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Proofread a page →
About 2 texts to rework per day (scientific English), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Translate a page →
About 2 pages to translate per week, assumption; share done with AI set high, translation being among the tasks most often delegated to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Generate a function →
About 1 analysis script per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Write an email →
About 12 emails per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : 10-slide presentation →
About 1 presentation per month (seminar, conference), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Conversation of 10 exchanges →
About 2 methodology discussions per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year →
Published observation on uses
Scope of the figure used. Generative AI applicability score for public relations specialists (code 27-3031), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Communications officer, 1 year → : Article of 1,000 words →
About 1 long article per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Conversation of 10 exchanges →
About 3 brainstorming sessions per week (angle, slogan, campaign), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Create an image →
About 5 supporting visuals per week for social media, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. server and data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Visual in 4 attempts →
About 2 campaign visuals searched for per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. images generated, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Translate a page →
About 2 pages to translate per week, assumption; share done with AI set high, translation being among the tasks most often delegated to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Web search with an AI →
About 3 web searches per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : 10-slide presentation →
About 2 presentations per month, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for HR officers (code 13-1071), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Human resources, 1 year → : Proofread a page →
About 2 texts to rework per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Write a page →
About 1 page per day (job description, letter, note), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Write meeting minutes →
About 4 interview reports per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Conversation of 10 exchanges →
About 2 delicate situations to prepare per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Web search with an AI →
About 2 web searches per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : 10-slide presentation →
About 1 presentation per month, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for accountants and auditors (code 13-2011), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Published observation on uses
Scope of the figure used. Generative AI applicability score for accounting clerks (code 43-3031), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Accountant, 1 year → : Write an email →
About 12 emails per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Proofread a page →
About 2 texts to rework per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Query a PDF →
About 2 long documents to query per week (tax file, contract, report), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Summarise 30 pages →
About 1 regulatory text or report to summarise per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Write meeting minutes →
About 1 set of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Web search with an AI →
About 2 web searches per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : 10-slide presentation →
About 1 presentation per month (closing, budget), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for software developers (code 15-1252), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Developer, 1 year → : Query a PDF →
About 3 long documentation sets queried per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Write an email →
About 8 long emails or messages per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Write meeting minutes →
About 2 minutes or incident reports per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Web search with an AI →
About 3 web searches per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for office employees (code 43-9061), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Published observation on uses
Scope of the figure used. Generative AI applicability score for interpreters and translators (code 27-3091), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Office employee, 1 year → : Write a page →
About 1 page written per day (note, letter, procedure), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Web search with an AI →
About 5 web searches per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Conversation of 10 exchanges →
About 3 reflection sessions per week (preparing a decision, unblocking a file), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Write meeting minutes →
About 2 sets of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Summarise 30 pages →
About 1 long document to read per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Translate a page →
About 1 page to translate per week, assumption; share done with AI set high, translation being among the tasks most often delegated to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : 10-slide presentation →
About 2 presentations per month, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for executive assistants (code 43-6011), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Executive assistant, 1 year → : Write a page →
About 3 pages per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Write meeting minutes →
About 4 sets of minutes per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : 10-slide presentation →
About 2 presentations to format per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Translate a page →
About 2 pages to translate per week, assumption; share done with AI set high, translation being among the tasks most often delegated to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Analyse a spreadsheet →
About 2 spreadsheets per week (budget, diary), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Summarise 30 pages →
About 1 long document to summarise per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Web search with an AI →
About 4 searches per day (travel, suppliers), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for secondary school teachers (code 25-2031), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Teacher, 1 year → : Write an email →
About 5 emails per day (families, colleagues), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Write meeting minutes →
About 1 report per week (class council, meeting), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Summarise 30 pages →
About 10 long texts to read per year (curriculum, report), assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Teacher, 1 year → : Translate a page →
About 1 page to translate every 2 weeks, assumption; high share done with AI, translation being among the tasks most often handed to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Generative AI applicability score for nurses (code 29-1141), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Published observation on uses
Scope of the figure used. Generative AI applicability score for electricians (code 47-2111), 200,000 conversations with Bing Copilot, from 0 to 1
Limit. Used for the occupation coefficient: square root of the score divided by 0.25, the score of office employees
Field occupation, 1 year → : Quick question →
About 4 questions per day (procedure, breakdown, instruction), assumption; share done with AI based on the Crédoc Baromètre du numérique (edition 2026, data 2025), adjusted to the occupation, low for occupations where the work is mainly physical (Microsoft, 2025).
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Write an email →
About 3 emails or messages per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Voice mode, 1 minute →
About 5 minutes of dictation per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. voice heard, voice spoken, context re-read, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Query a PDF →
About 2 long manuals or protocols to query per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Translate a page →
About 1 exchange to translate per week (patient, client, manual), assumption; high share done with AI, translation being among the tasks most often handed to AI (Microsoft, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Proofread a page →
About 2 texts to rework per week, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Field occupation, 1 year → : Web search with an AI →
About 1 web search per day, assumption; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. searches, pages read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Role and limits
Author. National Center for O*NET Development, US Department of Labor
What the publication covers. tasks of each occupation and frequency reported by job holders, in 7 classes from "once a year or less" to "every hour or more"
What it does not cover. occupations without a frequency survey, such as data scientists
Reading by the database. The reference classification of tasks; US job holders
Manager or executive, 1 year → : Summarise 30 pages →
About 3 reports to read per week, reviewing results being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Analyse a spreadsheet →
About 1 spreadsheet to analyse per day, analysing collected data being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Consultant, 1 year → : Summarise 30 pages →
About 3 long documents to summarise per week, information gathering being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, summary written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. lawyers: interpreting laws and decisions for clients, share who do so daily or more
Limit. Support for the frequency of the task in the basket; US job holders
Legal adviser, 1 year → : Review a contract →
About 4 contracts of 20 pages reviewed per week, text analysis being a daily task for most lawyers in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Quick question →
About 6 legal questions per day, interpreting texts being a daily task in O*NET; share done with AI based on the Crédoc Baromètre du numérique (2026 edition, 2025 data), adjusted to the occupation.
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Legal adviser, 1 year → : Conversation of 10 exchanges →
About 1 case strategy to prepare per week, preparing arguments being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Researcher, 1 year → : Analyse a spreadsheet →
About 3 datasets to analyse per week, data analysis being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Write a page →
About 2 pages per day (press release, publication, letter), updating content being a daily task in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Write an email →
About 20 emails per day, media requests being a daily task in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Write an email →
About 20 emails per day, explaining the rules to employees being a daily task in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Human resources, 1 year → : Query a PDF →
About 2 batches of applications to sort per week, application review being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. document read, response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. accountants and auditors: writing detailed assignment reports, share who do so daily or more (most often more than once a year)
Limit. Support for the frequency of the task in the basket; US job holders
Accountant, 1 year → : Analyse a spreadsheet →
About 2 spreadsheets to analyse per day (reconciliations, checks), the core of the occupation in O*NET; share done with AI set low, data analysis carrying little weight in messages (OpenAI, 2025), adjusted to the occupation.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Accountant, 1 year → : Write a page →
About 3 pages of notes or reports per week, report writing being mostly monthly or annual in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Write a page →
About 2 pages of documentation per week, documentation being weekly or more often in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. office employees: opening, sorting mail and answering correspondence, share who do it daily or more
Limit. Support for the frequency of the task in the basket; US job holders
Office employee, 1 year → : Write an email →
About 15 emails written per day, correspondence being a daily task for office workers in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Write an email →
About 30 emails per day, mail being handled every day in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Executive assistant, 1 year → : Proofread a page →
About 4 texts to rework per day, preparing letters being a daily task in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation using the Microsoft study of 200,000 conversations.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. secondary school teachers: teaching (lessons, discussions, demonstrations), share of permanent staff who do so daily or more often
Limit. Support for the frequency of the task in the basket; US job holders
Published observation on uses
Scope of the figure used. secondary school teachers: preparing, administering and marking assessments, share who do so daily or more often (78% weekly or more often)
Limit. Support for the frequency of the task in the basket; US job holders
Published observation on uses
Scope of the figure used. secondary school teachers: setting and marking exercises and homework, share who do so daily or more often
Limit. Support for the frequency of the task in the basket; US job holders
Teacher, 1 year → : Conversation of 10 exchanges →
About 1 sequence to adapt per week, adapting to pupils' needs being daily in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. nurses: keeping accurate records and reports, share who do so daily or more often
Limit. Support for the frequency of the task in the basket; US job holders
Field occupation, 1 year → : Write meeting minutes →
About 3 intervention reports or handovers per day, record keeping being several times a day for nurses in O*NET; share done with AI taken from the value common to all occupations for this task, then adjusted to the occupation from the Microsoft study of 200,000 conversations.
Scope of the figure used. simple tasks, data centre included
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
OpenAI · September 2025
How People Use ChatGPT (NBER Working Paper 34255) ↗ (new tab)
Role and limits
Author. Aaron Chatterji, Tom Cunningham, David Deming, Zoë Hitzig, Christopher Ong, Carl Shan, Kevin Wadman (OpenAI, Duke, Harvard)
What the publication covers. exact message volumes of the consumer plans; classification of a sample of about 1.1 million conversations by topic, intent and link to work
What it does not cover. enterprise plans, under-18s, logged-out users
Reading by the database. The only published measurement of the volumes and uses of a major assistant; internal data, automatic classification
Manager or executive, 1 year →
Published observation on uses
Scope of the figure used. Share of writing in the work messages of management and leadership occupations
Limit. Supporting evidence for the share of emails and texts in the manager's task basket
Consultant, 1 year → : Proofread a page →
About 4 texts to rework per day; share done with AI based on the share of work messages that edit a supplied text (OpenAI, 2025), adjusted to the occupation.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Communications officer, 1 year → : Proofread a page →
About 4 texts to rework per day; share done with AI based on the share of work messages that edit a supplied text (OpenAI, 2025), adjusted to the occupation.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Published observation on uses
Scope of the figure used. Share of programming in consumer ChatGPT messages
Limit. Code mostly goes through development tools and agents, outside this measurement
Published observation on uses
Scope of the figure used. Share of technical help in the work messages of computing occupations
Limit. Support for the place of technical questions in the developer's task basket
Published observation on uses
Scope of the figure used. Share of work-related messages, consumer plans, June 2025 (47% a year earlier)
Limit. Used to check the average across occupations
Published observation on uses
Scope of the figure used. Share of writing in work-related messages, July 2025
Limit. Support for the common grid: writing is the first use at work
Published observation on uses
Scope of the figure used. About two thirds of writing requests modify a text supplied (rework, critique, translation, summary)
Limit. Support for the high use for reworking, translating and summarising, lower for writing from scratch
Office employee, 1 year → : Proofread a page →
About 3 texts to rework per day; share done with AI based on the share of work messages that ask for a supplied text to be edited (OpenAI, 2025), adjusted to the occupation.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Office employee, 1 year → : Analyse a spreadsheet →
About 3 spreadsheets to analyse per week, assumption; share done with AI set low, data analysis carrying little weight in messages (OpenAI, 2025), adjusted to the occupation.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
OpenAI · December 2025
Role and limits
What the publication covers. actual use by ChatGPT enterprise customers and a survey of 9,000 employees
What it does not cover. the consumer plans and any absolute volume per employee
Reading by the database. Publisher report; the gaps between users are the useful data
Communications officer, 1 year →
Published observation on uses
Scope of the figure used. Gap between the 95th percentile and the median for writing and communication
Limit. Supporting evidence for the gaps between levels for writing
Published observation on uses
Scope of the figure used. Gap between the 95th percentile and the median for code messages, the largest of all uses
Limit. Support for the high intensive use for code
Published observation on uses
Scope of the figure used. Messages of the most active employees (95th percentile) relative to the median employee, enterprise customers
Limit. Support for the gaps between the light, typical and intensive levels
Role and limits
What the publication covers. audio tokens: one per 100 milliseconds of the user's speech, one per 50 milliseconds of the assistant's speech
What it does not cover. electricity
Reading by the database. The publisher's billing rule, read in the page excerpt
Voice mode, 1 minute → : Read 1,000 tokens →
What the user says, half of the minute, in audio tokens
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Voice mode, 1 minute → : Write 1,000 tokens →
What the AI answers, the other half, in audio tokens
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Voice mode, 1 minute → : Read 1,000 tokens →
Length convention: one minute of voice heard by an AI
Scope of the figure used. One audio token per 100 milliseconds of the user's speech at OpenAI; Gemini counts 32 per second, or 1,920 per minute
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Voice mode, 1 minute → : Write 1,000 tokens →
Length convention: one minute of voice spoken by an AI
Scope of the figure used. One audio token per 50 milliseconds of the assistant's speech at OpenAI
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Projet Celsius · November 2026
Tokens per word, per line of code and per cell, measured with 7 public tokenisers
Role and limits
What the publication covers. 479,888 words from 196 blog articles and their English translation, code from the site and the Python library, a table of measurements; tokenisers of GPT-4o, GPT-4, Llama 3, Gemini (via Gemma 3), Mistral, DeepSeek and Qwen
What it does not cover. the Claude tokeniser, not public
Reading by the database. A measurement on technical texts, figures included; to be repeated if publishers change their tokeniser
Projet Celsius measurement, repeated on request with the public tokenisers
Proofread a page → : Read 1,000 tokens →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Proofread a page → : Write, with reasoning →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Summarise 30 pages → : Read 1,000 tokens →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Summarise 30 pages → : Write, with reasoning →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Conversation of 10 exchanges → : Re-read from cache →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Write a page → : Write, with reasoning →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Analyse a spreadsheet → : Read 1,000 tokens →
The spreadsheet, 200 rows and 10 columns, and the question
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Analyse a spreadsheet → : Read 1,000 tokens →
Length convention: a spreadsheet of 200 rows
Scope of the figure used. A table of 200 rows and 10 columns of numbers and short labels: 4.7 tokens per cell, measured on a real table
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Deep research → : Write 1,000 tokens →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Query a PDF → : Read 1,000 tokens →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Translate a page → : Write, with reasoning →
The translated page, counted as a French page
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Translate a page → : Read 1,000 tokens →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Translate a page → : Write, with reasoning →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Code suggestion → : Read 1,000 tokens →
Length convention: a line of code
Scope of the figure used. From 10 tokens per line (Python standard library) to 19 (the TypeScript code of this site), measured with GPT-4o tokenisation
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Code suggestion → : Write 1,000 tokens →
Length convention: a line of code
Scope of the figure used. From 10 tokens per line (Python standard library) to 19 (the TypeScript code of this site), measured with GPT-4o tokenisation
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Get code reviewed → : Write, with reasoning →
Length convention: one page
Scope of the figure used. A report page, headings and paragraphs included: between the 1,500-character sheet (250 words) and a full A4 page in single spacing (about 550 words, at 6.1 characters per word measured on our articles)
Limit. This publication supports the quantity of text or content processed, not its carbon footprint.
Stack Overflow · 2025
Role and limits
What the publication covers. frequency of use of AI and agents, tasks handed over, about 33,000 respondents
What it does not cover. the number of requests
Reading by the database. Online survey of volunteer developers, more engaged than average
Published observation on uses
Scope of the figure used. Use AI tools every day in their development work (17.4% every week), 26,004 respondents
Limit. Support for the developer's typical use
Published observation on uses
Scope of the figure used. Write code partly with AI (16.9% mostly with AI, on a different base of respondents: the two do not add up)
Limit. Support for the use of function generation
Published observation on uses
Scope of the figure used. Use AI agents at work every day (9.2% every week)
Limit. Support for the use of agents in 2025
Developer, 1 year → : Code suggestion →
About 200 suggestions offered per day when the assistant is active, assumption; share done with AI based on the half of developers who use AI every day (Stack Overflow, 2025).
Scope of the figure used. code read, suggestion written, small model, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Generate a function →
About 6 functions written per day, assumption; share done with AI based on the 59% of respondents who write code partly with AI (Stack Overflow, 2025); replaced by agent hours when there are agents.
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Get code reviewed →
About 2 code reviews per day, assumption; share done with AI based on the 23% of respondents who have their code partly reviewed by AI (Stack Overflow, 2025).
Scope of the figure used. tokens read and written, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Quick question →
About 10 technical questions per day, assumption; share done with AI set high, looking for an answer being the main stated use (Stack Overflow, 2025).
Scope of the figure used. computation of the response, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Developer, 1 year → : Conversation of 10 exchanges →
About 2 debugging sessions per day, fixing code being weekly or more often in O*NET; share done with AI based on the 47% of respondents who debug partly with AI (Stack Overflow, 2025).
Scope of the figure used. messages, context re-read from cache, responses, data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Stack Overflow · 2026
Role and limits
What the publication covers. daily duration of AI use, code assistants and agents, tasks handed over
What it does not cover. the share of professional developers
Reading by the database. Same survey a year later; some questions changed
Published observation on uses
Scope of the figure used. Use AI more than 4 hours per day (24% for 2 to 4 hours, 25% for 1 to 2 hours), 10,926 respondents
Limit. Support for agent hours: the median falls between 2 and 4 hours per day
Published observation on uses
Scope of the figure used. Use code assistants and agents every day
Limit. Support for agent tooling in 2026
Zhang and Nam · June 2026
HANSEL: Extracting Breadcrumbs from Web Agent Trajectories for Interactive Verification ↗ (new tab)
Role and limits
Author. Yujin Zhang and Daye Nam
What the publication covers. 45 web agent trajectories: 13.16 steps and 6.02 pages visited per task on average
What it does not cover. electricity
Reading by the database. A small sample, which gives the order of magnitude of an agent's steps
Web search with an AI → : Read 1,000 tokens →
The 5 pages found, read by the model, and the question: web agents visit 6 pages per task on average
Scope of the figure used. server, excluding the data centre
Limit. The quantity and the share entrusted to AI are database assumptions, not a measured footprint.
Energy and factors
The references of the selected values and of the entries used in this calculation.
Delavande et al. · September 2025
Role and limits : Delavande et al.
Author. Delavande, Pierrard and Luccioni
What the publication covers. GPU, processor, memory
What it does not cover. data centre
Reading by the database. Seven open models
5-second clip → : Second of video →
The seconds of the measured video, 5.4
Scope of the figure used. GPU, processor, memory; excluding the data centre
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Role and limits : Google
What the publication covers. chips, host, standby machines, data centre, on-site water
What it does not cover. manufacturing, training
Reading by the database. The only production measurement published by a provider
Reference of the selected value
Scope of the figure used. Median of a text prompt; chips, host, standby machines, data centre
Limit. The only production measurement published by a provider, not audited by a third party
Google · January 2009
Role and limits : Google
What the publication covers. a search, data centre
Reading by the database. Never remeasured
Reference of the selected value
Scope of the figure used. A search, data centre included
Limit. Never publicly remeasured; Google's AI answers are not published
Luccioni et al. · October 2024
Role and limits : Luccioni et al.
Author. Luccioni, Jernite and Strubell
What the publication covers. GPU, processor, memory
What it does not cover. data centre
Reading by the database. Open models only
Create an image → : Generate an image →
One image, of the chosen size: standard by default
Scope of the figure used. GPU, processor, memory; excluding the data centre
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Reference for the variant "Standard image, from 512 to 1,024 pixels per side"
Scope of the figure used. Average of the open models tested, 1,000 images; median 1.35 Wh, maximum 11.49 Wh
Limit. Open models only; closed generators publish nothing
Microsoft Research · 2026
Estimation de l'énergie des requêtes en production (Joule) ↗ (new tab)
Role and limits : Microsoft Research
Author. Microsoft Research, Oviedo et al.
What the publication covers. H100 nodes, production
Reading by the database. Calibrated on production
Reference of the selected value
Scope of the figure used. Median, 500 tokens read and 300 written, H100 nodes
Limit. Not a measurement, but calibrated on production; judges public estimates to be 4 to 20 times too high
The 3 other publications in this section
ML.ENERGY · February 2026
Role and limits
Author. Chung et al., University of Michigan
What the publication covers. GPU energy per token, per response, per image, in steady state, according to the number of requests served at once
What it does not cover. processors, host memory, data centre cooling, manufacturing
Reading by the database. The best open benchmark: it serves to compare model sizes with each other, not as a production value
Support for the "Small model" setting
Scope of the figure used. A model of about 8 billion active parameters, like Llama 3.1 8B: about 6 times less electricity per token than the medium model, according to the ML.ENERGY benchmark.
Limit. The setting is a convention of the database supported by this publication.
Reference of the variant "Small model, about 8 billion parameters"
Scope of the figure used. Llama 3.1 8B, one H100 GPU, 64 requests at a time, GPU only, per 1,000 tokens written, reading of the question included
Limit. GPU only, on a benchmark: the database keeps only the ratio to the 70 billion model, 1 to 6.7
The measurement covers a different scope; it is not a direct measurement of this entry.
OpenAI · June 2025
Role and limits
Author. Sam Altman, OpenAI
What the publication covers. an average query
What it does not cover. method, scope
Reading by the database. One sentence, no method
Reference of the selected value
Scope of the figure used. "An average query", with no published method
Limit. No published scope; consistent with Google
Watershed · Aug. 2026
Estimating GHG Emissions from AI Use: Framework for Corporate-Level Measurement ↗ (new tab)
Role and limits
Author. Bistline, Ulissi, Davis, Glidden, Joyce, Li, Mohsenin and Suh (Watershed)
What the publication covers. factors per 1,000 tokens read and written at the server before PUE (host and standby machines included), PUE separate (1.10 for a large operator, 1.56 for an unknown operator)
What it does not cover. clean measurements, the model and the hardware
Reading by the database. Default factors proposed for review (public data as of June 2026); for a large operator, taken from Google's measurement published in 2025 (Elsworth et al.), which already contains the data centre overhead; constant per token, they underestimate long contexts and agents according to the authors
Reference for the variant "Medium model, Watershed default factor"
Scope of the figure used. Default factor for 1,000 tokens read, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
Limit. A default value that Watershed derives from Google's measurement (0.24 Wh for the median text query of the Gemini app), with about 500 tokens per query and a written token costing 3 times a read token; Watershed does not give the split between read and written tokens. The 0.24 Wh already include the overhead of Google's data centre (0.02 Wh, 8%): with 375 tokens read and 125 written, the simplest reading, the factor contains it, and the PUE of 1.10 applied afterwards counts it a second time (about 9%). Values proposed for review, final at the next version of the framework
Reference for the variant "Medium model, Watershed default factor"
Scope of the figure used. Default factor for 1,000 tokens written, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
Limit. 3 times the reading factor, based on the ratio between writing and reading selected by Watershed (Patel et al., Splitwise, 2024). Same derivation and same caveat as for reading: Google's 0.24 Wh already include about 8% of data centre. Values proposed for review, final at the next version of the framework
Write an email → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Reference of the selected value
Scope of the figure used. Default factor for a large operator, from the server to the data centre; 1.56 for an unknown operator
Limit. A default value, not a measurement
Quick question → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Proofread a page → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Summarise 30 pages → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Conversation of 10 exchanges → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Write a page → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Analyse a spreadsheet → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Create an image → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Deep research → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Query a PDF → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Web search with an AI → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Article of 1,000 words → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Translate a page → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Generate a function → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Visual in 4 attempts → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
5-second clip → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Reference factor: Medium model, Watershed default factor
Scope of the figure used. Default factor for 1,000 tokens read, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
Limit. This measurement concerns the reference variant, not a measurement of "Small model, about 8 billion parameters". A default value that Watershed derives from Google's measurement (0.24 Wh for the median text query of the Gemini app), with about 500 tokens per query and a written token costing 3 times a read token; Watershed does not give the split between read and written. The 0.24 Wh already include the overhead of Google's data centre (0.02 Wh, 8%): with 375 tokens read and 125 written, on the simplest reading, the factor contains it, and the PUE of 1.10 applied afterwards counts it a second time (about 9%). Values proposed for review, to be finalised in the next version of the framework
Reference factor: Medium model, Watershed default factor
Scope of the figure used. Default factor for 1,000 tokens written, medium model, large operator, at the server before PUE (Watershed table 2, public data as of June 2026)
Limit. This measurement concerns the reference variant, not a measurement of "Small model, about 8 billion parameters". 3 times the read factor, based on the write-to-read ratio adopted by Watershed (Patel et al., Splitwise, 2024). Same derivation and same caveat as for reading: Google's 0.24 Wh already include about 8% for the data centre. Values proposed for review, to be finalised in the next version of the framework
Code suggestion → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Get code reviewed → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Voice mode, 1 minute → : Data centre PUE →
The data centre, convention C3
Scope of the figure used. building cooling and losses
Limit. This figure covers part of the calculation; it does not measure the whole entry.
Conversion into carbon
The source of the electricity factor of the selected location.
Worldwide: 458 gCO2e/kWh, life cycle.
Role and limits : Ember
Author. Ember, Our World in Data
What the publication covers. electricity generation over the life cycle, by country
What it does not cover. guarantees of origin
Reading by the database. The same source for all countries
Conversion of electricity into carbon: worldwide
Scope of the figure used. the production of electricity over the life cycle, including power plant construction and fuels
Limit. An average for the chosen country, not the electricity contract or the data centre of a specific service.
Nature of the data
Published readings, conventions, assumptions
- Projet Celsius assumptions 100 %
Calculation assumptions
Quantities, nature and source of each assumption
Manager or executive, 1 year : × 55,000
Executives and managers in the civil service: 110,000; AI use 2 times that of the average worker
Consultant, 1 year : × 100,000
Engineers, consultants and research officers in the civil service: 200,000; AI use 2 times that of the average worker
Lawyer, 1 year : × 25,000
Lawyers in the civil service: 50,000; AI use 2 times that of the average worker
Researcher, 1 year : × 50,000
Researchers in the civil service: 100,000; AI use 2 times that of the average worker
Communications officer, 1 year : × 15,000
Communication and marketing officers in the civil service: 30,000; AI use 2 times that of the average worker
Human resources, 1 year : × 40,000
Human resources officers in the civil service: 80,000; AI use 2 times that of the average worker
Accountant, 1 year : × 40,000
Accountants and finance professionals in the civil service: 100,000; AI use 1.6 times that of the average worker
Developer, 1 year : × 43,800
Developers in the civil service: 70,000; AI use 2.5 times that of the average worker
Office worker, 1 year : × 319,000
Office employees and technicians in the civil service: 850,000; AI use 1.5 times that of the average worker
Executive assistant, 1 year : × 56,300
Assistants and secretaries in the civil service: 150,000; AI use 1.5 times that of the average worker
Teacher, 1 year : × 163,000
Teachers in the civil service: 1.00 million; AI use 0.65 times that of the average worker
Field-based occupation, 1 year : depending on the settings
Skilled field occupations (care workers, technicians, social work) in the civil service: 1.30 million; AI use 0.65 times that of the average worker
Field-based occupation, 1 year : depending on the settings
Service and sales occupations in the civil service: 950,000; AI use 0.45 times that of the average worker
Field-based occupation, 1 year : depending on the settings
Manual workers, farmers and manual occupations in the civil service: 860,900; AI use 0.27 times that of the average worker
Methodology conventions
With their alternatives
| The default location | Without other information, the servers are in the world, at the world average electricity |
|---|---|
| Electricity, location-based, on a life cycle basis | The kilowatt-hour is that of the servers' country, on a life cycle basis, from a single source for all countries |
| Two significant figures | A calculated value is rounded to two significant figures, a published value is reproduced with the precision of its source; the calculation keeps full precision |
| The formula | the sum of what makes up the entry, then CO2e: electricity (Wh) multiplied by the carbon intensity of the server location (gCO2e/kWh), divided by 1,000 |
| Calculate a case without an entry | The formula, the default values and a step-by-step example |
History
Entry in version 0.1 of 2 November 2026, in the database in version 0.4
- 2 November 2026Version 0.1 of the entry, as of the date of its current value.
- 31 October 2026Creation of the database, in its first version (0.1).
