Discover AbbeyCambridge
How much of the work could AI tools speed up?
This is about the tasks in the job, not whether there will be jobs. It does not predict job losses. It describes what generally available AI tools could help with today, based on the jobs residents do. It is not a forecast.
Residents in work in Cambridge (70,595 people)
- A small part 39.4%
- Some 23.2%
- Much 35.2%
- Not yet assessed 2.2%
England and Wales for comparison
- A small part 50.8%
- Some 19.4%
- Much 28.5%
- Not yet assessed 1.3%
Small, some and much are always shown together because no single score describes a job. They add up to 100%. "Not yet assessed" means we have no reliable rating for those occupations; it does not mean low or high. At least 24.3% of residents work in occupation groups where every assessed career is rated "much" (20% in England and Wales).
CareerMash staging data (11 Oct 2026), approved by the owner for this preview, subject to change; not yet independently validated
How do we know this?
Occupation-adjusted estimate, residents, Census 2021 occupations, national mix within each group. Local share of employed residents in each of 104 occupation groups (SOC 2020 three-digit) multiplied by the national mix of CareerMash categories inside that group, weighted by UK employment (ONS APS, April 2025 to March 2026). Small, some, much and not assessed add to 100 percent and are always shown together. The floor is the share of residents in occupation groups where every assessed career is rated "much".
Period: Census 2021 (21 March 2021). Quality: medium.
Limits
- Resident basis: this counts people where they live, not where they work. Some residents work in other areas.
- Census occupations are as at 21 March 2021, a COVID-19 period. The CareerMash categories are task evidence dated 11 October 2026. This is not a description of current employment.
- Each local figure is an estimate: local occupation shares multiplied by a national mix inside each occupation group. It does not use local employers or local jobs, and a place can differ from the national mix inside a group.
- Occupation groups are weighted by UK employment from the ONS Annual Population Survey (April 2025 to March 2026), a survey with sampling variability. The 2021 Census and that survey count some occupations differently, so the UK survey profile and the England and Wales Census profile are not the same; compare a place with England and Wales on this page, not with UK survey figures.
- Within a SOC unit group the category mix counts each assessed CareerMash career equally; the data holds no per-career employment. 1,547 of 1,840 careers are assessed (293 are withheld and never used). Occupations with no assessed career are shown as "not assessed".
- The exposure categories describe how much of a job's tasks generally available AI tools could speed up. They are not job losses, adoption, robotics or physical automation, and not an outlook for any occupation or place.
- Time shares and categories were estimated by Claude models and have not been independently validated by another model family. CareerMash staging data (11 Oct 2026), subject to change.
- Robot use is shown only for industries where a published figure exists (Eurostat, EU firms). Robotics exposure of occupations, and science and market effects, are not yet assessed.
Where it varies: by type of occupation
The same mix, split by the nine broad groups of occupation. Groups differ a lot: the local mix depends on which occupations are common here.
- A small part
- Some
- Much
- Not yet assessed
- Managers, directors and senior officials10.5% of residents in work (12.8% in England and Wales)65.6%30.9%
- Professional occupations42.4% of residents in work (20.2% in England and Wales)19%28.1%48%
- Associate professional and technical occupations12.1% of residents in work (13.2% in England and Wales)32%20%47.2%
- Administrative and secretarial occupations5.8% of residents in work (9.3% in England and Wales)8.3%90%
- Skilled trades occupations5.1% of residents in work (10.3% in England and Wales)92%
- Caring, leisure and other service occupations6.8% of residents in work (9.4% in England and Wales)92.8%
- Sales and customer service occupations4.9% of residents in work (7.5% in England and Wales)82.1%10.9%
- Process, plant and machine operatives3.4% of residents in work (7% in England and Wales)96.9%
- Elementary occupations9% of residents in work (10.5% in England and Wales)97.3%
CareerMash staging data (11 Oct 2026), approved by the owner for this preview, subject to change; not yet independently validated
What still needs a person
Alongside exposure, each occupation is rated on whether the work itself depends on a person. These are separate questions: work can be speeded up by tools and still need someone present, trusted or accountable.
This place England and Wales
Share of employed residents. "Done in person" is about presence, not about robots: it is not a measure of physical automation. From OpenAI's AI Jobs Transition Framework (2026), matched to careers by Careermash. UK statutory-register rules and recorded Careermash overrides also apply.
CareerMash staging data (11 Oct 2026), approved by the owner for this preview, subject to change; not yet independently validated
How do we know this?
The same occupation-adjusted aggregation as the exposure estimate, applied to the human necessity class of each assessed career. This is a separate dimension from exposure. "Physical" means much of the work has to be done in person; it is not a measure of robotics or physical automation.
- Shares are of employed residents in occupations, built the same way as the exposure estimate, with the same resident, 2021 and estimate caveats.
- "Physical" means much of the work has to be done in person, with your hands or on site. It is a human necessity classification, not a robotics exposure.
- Some assessed careers are "Not yet classified" and occupations with no assessed career are not classified at all. Both are shown.
- CareerMash staging data (11 Oct 2026), subject to change.
Local industries
What residents in work do for a living, by industry, compared with England and Wales. This is where people live, not where businesses are.
This place England and Wales
Census 2021, industry of the main job of residents in work (ten largest of 18 groups shown).
Robots, science and market effects
These describe firms and industries, not individual jobs. They are not forecasts and do not say anything about job losses.
Three further questions are kept apart from AI exposure. Only the first has a published source that can be tied to industries, so only the first is filled in.
Robots: how common is their use among firms in the industry?
Percent of European Union firms with 10 or more employees that use AI-enabled autonomous robots, 2025. All industries covered: 1.4%. This counts how common it is for firms in the industry to use autonomous robots. It says something about how far robotics has reached the industry, not about how many jobs or tasks are affected. It covers a narrow type of robot, so it understates the use of conventional industrial robots, which the survey reports only for all industries combined.
No published figure for: Education; Human health and social work activities; Other (arts, entertainment and recreation; other services; households as employers; extraterritorial bodies); Public administration and defence; Financial and insurance activities; Agriculture, Forestry and fishing; Mining and quarrying. The survey does not cover them, and nothing is inferred.
Science and medical advances
Market and demand
Source and licence
Eurostat: Use of artificial intelligence technologies in enterprises by NACE Rev. 2 activity (isoc_eb_ain2), indicator E_AI_TAR; Community survey on ICT usage and e-commerce in enterprises. https://ec.europa.eu/eurostat/databrowser/view/isoc_eb_ain2
Licence: Eurostat reuse policy (CC BY 4.0). Retrieved 2026-10-12. NACE Rev. 2 and UK SIC 2007 share the same section letters. The survey does not cover agriculture, mining, financial services, public administration, education, health and social work, or arts and other services. Context figures: Eurostat isoc_eb_p3d (2020) and Eurostat news, 21 January 2019.
Local circumstances that shape adaptation
Cambridge figure
Dimensions are shown separately with no composite score. They describe local circumstances in March 2021 and are not a judgement of residents.
Qualifications of residents aged 16+
This place England and Wales
What residents aged 16+ were doing on Census day
This place England and Wales
Census day fell during the COVID-19 pandemic, when many people were on furlough.
12.5% of people in employment are self-employed (England and Wales 16.6%). This is a share of people in work, not of all residents, so it is not comparable with the bars above.
How widely employment is spread
The effective number of sectors is 1 divided by the sum of squared sector shares across 18 industry groups. A higher number means employment is spread across more sectors; a lower number means it is concentrated in fewer. Neither is good or bad in itself.
Census 2021, 21 March 2021, during the COVID-19 pandemic. These describe local circumstances, not people's ability to adapt.
CareerMash staging data (11 Oct 2026), approved by the owner for this preview, subject to change; not yet independently validated
Method and limits
Occupation-adjusted estimate, residents, Census 2021 occupations, national mix within each group. Local share of employed residents in each of 104 occupation groups (SOC 2020 three-digit) multiplied by the national mix of CareerMash categories inside that group, weighted by UK employment (ONS APS, April 2025 to March 2026). Small, some, much and not assessed add to 100 percent and are always shown together. The floor is the share of residents in occupation groups where every assessed career is rated "much".
- Resident basis: this counts people where they live, not where they work. Some residents work in other areas.
- Census occupations are as at 21 March 2021, a COVID-19 period. The CareerMash categories are task evidence dated 11 October 2026. This is not a description of current employment.
- Each local figure is an estimate: local occupation shares multiplied by a national mix inside each occupation group. It does not use local employers or local jobs, and a place can differ from the national mix inside a group.
- Occupation groups are weighted by UK employment from the ONS Annual Population Survey (April 2025 to March 2026), a survey with sampling variability. The 2021 Census and that survey count some occupations differently, so the UK survey profile and the England and Wales Census profile are not the same; compare a place with England and Wales on this page, not with UK survey figures.
- Within a SOC unit group the category mix counts each assessed CareerMash career equally; the data holds no per-career employment. 1,547 of 1,840 careers are assessed (293 are withheld and never used). Occupations with no assessed career are shown as "not assessed".
- The exposure categories describe how much of a job's tasks generally available AI tools could speed up. They are not job losses, adoption, robotics or physical automation, and not an outlook for any occupation or place.
- Time shares and categories were estimated by Claude models and have not been independently validated by another model family. CareerMash staging data (11 Oct 2026), subject to change.
- Robot use is shown only for industries where a published figure exists (Eurostat, EU firms). Robotics exposure of occupations, and science and market effects, are not yet assessed.
- This section describes how occupations and industries of residents compare, using Census 2021 and CareerMash staging data. It does not forecast jobs, job losses, adoption of technology or the future of any place.
- Not shown, by design: any numeric score for an occupation or place, rankings of places, any figure for withheld careers, robotics exposure of occupations, science or market effects, and task text.
- Market effects and science effects are shown only where a cited, documented mechanism has been loaded. None is loaded in this version, so they read "not yet assessed". Robot use by industry comes from Eurostat and is added by exposure_ext.js.
CareerMash staging data (11 Oct 2026), approved by the owner for this preview, subject to change; not yet independently validated
Sources used on this page
9- Office for National Statistics: Annual Population Survey ad hoc 3501: 4-digit SOC 2020 by sex, April 2025 to March 2026Period: April 2025 to March 2026 · Open Government Licence v3.0
- CareerMash: AI exposure and human necessity, release exposure-v2.0 (staging)Period: Evidence date 11 October 2026 · Owner-approved for this preview build; ONS SOC 2020 task lists OGL v3, O*Net 31.0 CC BY 4.0
- Office for National Statistics: Census 2021 TS060 Industry (SIC 2007 sections and divisions)Period: 21 March 2021 · Open Government Licence v3.0
- Office for National Statistics: Census 2021 TS063 Occupation (SOC 2020 major groups)Period: 21 March 2021 · Open Government Licence v3.0
- Office for National Statistics: Census 2021 TS064 Occupation - minor groups (SOC 2020, usual residents aged 16+ in employment)Period: 21 March 2021 · Open Government Licence v3.0
- Office for National Statistics: Census 2021 TS066 Economic activity statusPeriod: 21 March 2021 · Open Government Licence v3.0
- Office for National Statistics: Census 2021 TS067 Highest level of qualificationPeriod: 21 March 2021 · Open Government Licence v3.0
- OpenAI: AI Jobs Transition Framework (human necessity categories), matched to careers by CareerMashPeriod: 2026 · CC BY 4.0, attribution required
- Office for National Statistics: Standard Occupational Classification 2020 unit group titlesPeriod: SOC 2020 · Open Government Licence v3.0
Registered offices are not operating locations. Ranks and deciles are never averaged. Census 2021 figures are labelled 2021.