Updated U.S. outlook through 2035. Compare net job growth with percentage growth; neither is a count of current vacancies.
Read the source notes ↓Check the argument.
The finding, the method, the limit. This is the reviewed evidence behind our briefings—not the automatically collected reading feed.
ReviewedWhat is newer—and what it changes.
Workplace app activity increased among adopters, using 2024 data. Activity is not output quality or business value.
Read the source notes ↓Technical workers report benefits. Read this survey alongside METR’s February warning about selection bias in its follow-up experiment.
Read the source notes ↓Danish work changed without large average earnings or hours effects in the early period. This adds context to U.S. hiring signals.
Read the source notes ↓New publication does not always mean new data. We retain older studies when they remain useful, and show newer follow-ups that qualify them. This is a selected review, not an exhaustive or continuously updated literature survey.
International AI Safety Report 2026
International scientific assessment · chaired by Yoshua Bengio
Finding. Future capability trajectories and employment effects remain uncertain; deployment and institutions matter alongside technical progress.
Limit. Not a forecast of an AGI date, an endorsement of a particular policy, or a test of which careers survive superintelligence.
Study design, period & location in source
- Period covered
- Evidence available for the February 2026 report; not a September tool evaluation
- Scope & method
- International expert review of general-purpose AI capabilities, emerging risks and risk management.
- Where to look
- Sections 1.3 (Capabilities by 2030), 2.3.1 (Labour market impacts) and the report scope statement.
Position: Levels of AGI for Operationalizing Progress on the Path to AGI
Morris and colleagues · Google DeepMind · ICML 2024
Finding. A system’s skill level, generality and deployment autonomy are different questions.
Limit. One proposed taxonomy, not a universal AGI definition. Its examples do not classify September 2026 systems or predict labor demand.
Study design, period & location in source
- Period covered
- 2023–2024 definitions and examples
- Scope & method
- Proposes separate dimensions for breadth of capability, performance and autonomy.
- Where to look
- Sections 3–4 and Table 1; Section 6 on autonomy and human–AI interaction.
Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile
NIST · AI 600-1
Finding. Identifies risks including confidently false output, data privacy and problematic human–AI configurations.
Limit. Guidance, not a career-outcomes experiment or assurance that a reviewed output is safe. This site does not certify compliance.
Study design, period & location in source
- Period covered
- 2024 cross-sector profile
- Scope & method
- Risks and suggested actions for organizations developing, evaluating or using generative AI.
- Where to look
- Section 2, especially 2.2 (Confabulation), 2.4 (Data Privacy), 2.7 (Human-AI Configuration); Section 3 actions.
Generative AI and Jobs: A Refined Global Index of Occupational Exposure
International Labour Organization · Working Paper 140
Finding. About 25% of employment is in occupations with some exposure; 3.3% is in the highest band.
Limit. Exposure describes potential task change, not observed adoption or job losses. The highest band is included in the 25%.
Study design, period & location in source
- Period covered
- 2025 task-capability assessment
- Scope & method
- Global employment; occupational tasks assessed using worker input, experts and AI, then linked to employment data.
- Where to look
- Executive summary; global employment by exposure gradient. DOI: 10.54394/HETP0387.
Workers’ exposure to AI: What indicators tell us—and what they don’t
International Labour Organization
Finding. Capability indicators alone cannot establish displacement, productivity or reskilling needs.
Limit. Methodological guidance, not a new estimate of employment change.
Study design, period & location in source
- Period covered
- 2026 review of exposure indicators
- Scope & method
- Review of what task-based AI exposure measures can establish.
- Where to look
- Discussion of indicators, adoption constraints and labor-market outcomes.
Labor market impacts of AI: A new measure and early evidence
Massenkoff & McCrory · Anthropic
Finding. No systematic rise in unemployment in more-exposed occupations; suggestive evidence of slower hiring of younger workers.
Limit. Association is not causation. One vendor’s usage is not all AI use, and aggregate stability can hide uneven effects.
Study design, period & location in source
- Period covered
- U.S. labor-market trends since late 2022
- Scope & method
- U.S. occupations, with exposure informed by observed Claude usage.
- Where to look
- Results on unemployment and hiring among young workers.
Still Waters, Rapid Currents: Early Labor Market Transformation under Generative AI
Humlum & Vestergaard · NBER Working Paper 33777
Finding. Task reorganization accompanied near-zero average effects on earnings and recorded hours; estimates rule out effects larger than 2%.
Limit. Early Danish outcomes, not a global or long-run prediction. Working paper, not a randomized trial.
Study design, period & location in source
- Period covered
- First two years after ChatGPT’s launch
- Scope & method
- Danish adoption surveys linked to administrative records; difference-in-differences analysis.
- Where to look
- Revised abstract and labor-market outcome estimates. DOI: 10.3386/w33777.
Generative AI at Work
Brynjolfsson, Li & Raymond · Quarterly Journal of Economics
Finding. 15% more issues resolved per hour on average; larger gains among less-experienced and lower-skilled agents.
Limit. Not randomized. Results vary by experience; experienced agents had small quality declines. One setting cannot establish economy-wide gains.
Study design, period & location in source
- Period covered
- Earlier-generation assistant; not a 2026 tool evaluation
- Scope & method
- Staggered rollout to 5,172 customer-support agents at one business-software firm.
- Where to look
- Abstract and heterogeneous effects; QJE 140(2), pages 889–942.
The Effects of Generative AI on High-Skilled Work: Evidence from Three Field Experiments with Software Developers
Cui and colleagues · Management Science
Finding. Pooled estimate: 26.08% more completed tasks (standard error 10.3 percentage points).
Limit. Individual trials were noisy. Output is not the same as quality or time saved. Includes Microsoft-affiliated authors and company-run trials.
Study design, period & location in source
- Period covered
- Earlier code-completion deployments; published in 2026
- Scope & method
- 4,867 developers across Microsoft, Accenture and a Fortune 100 company; randomized access to a coding assistant.
- Where to look
- Abstract; pooled results and online appendices.
Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity
METR
Finding. AI-allowed tasks took 19% longer; the reported confidence interval ran from 2% to 39% longer.
Limit. Small, specialized sample using early-2025 tools. Read alongside METR’s 2026 follow-ups, not as a current universal slowdown.
Study design, period & location in source
- Period covered
- February–June 2025 tools and tasks
- Scope & method
- 16 experienced open-source developers; 246 tasks in repositories they knew well.
- Where to look
- Main results and linked study paper.
We are Changing our Developer Productivity Experiment Design
METR
Finding. Raw estimates shifted toward speedups, but METR judged them unreliable as measures of the current effect.
Limit. Developers and tasks selected out of non-AI conditions; lower pay and concurrent agents complicated measurement.
Study design, period & location in source
- Period covered
- Experiment begun August 2025
- Scope & method
- Follow-up with returning and newly recruited open-source developers.
- Where to look
- Opening results and ‘Wider adoption of AI has made it more difficult to measure task-level productivity’.
Measuring the Self-Reported Impact of Early-2026 AI on Technical Worker Productivity
METR
Finding. Median perceived value multipliers ranged from 1.4× to 2× across differently worded questions.
Limit. Self-reports are not measured productivity; low response rates and inconsistent answers limit interpretation.
Study design, period & location in source
- Period covered
- Perceptions around March 2026
- Scope & method
- Convenience sample of 349 technical workers, including METR staff and professional networks.
- Where to look
- Methodology and ‘Productivity gains around March 2026’.
Adoption of Generative AI in the Workplace: Increasing and Shifting the Balance of Productivity and Communication Activity
Yu and colleagues · arXiv; Microsoft-affiliated research
Finding. Among higher-use adopters, productivity-app actions rose 21.2% and communication-app actions 7.1%.
Limit. App actions are not business value or output quality. Adoption was not randomized; vendor-affiliated preprint, not established consensus.
Study design, period & location in source
- Period covered
- January–September 2024 data
- Scope & method
- Microsoft 365 activity in 11 companies; difference-in-differences analysis of Copilot adoption.
- Where to look
- Abstract; Data and Methodology. Higher-use group: more than 100 uses over 20 weeks.
Fastest growing occupations, 2025–35
U.S. Bureau of Labor Statistics
Finding. Percentage growth and net new jobs highlight different opportunities; care occupations combine several large increases.
Limit. Not current vacancies, annual openings or an AI-immunity ranking. National medians do not establish local pay or access.
Study design, period & location in source
- Period covered
- 2025 baseline → 2035 projection; 2025 wages
- Scope & method
- U.S. national employment projections by detailed occupation.
- Where to look
- Table 1.3: employment change, percent and numeric. Employment figures in thousands are converted to jobs in our chart.
Employment projections, 2025–35
U.S. Bureau of Labor Statistics
Finding. Total employment is projected to grow 3.5%; healthcare support grows 13.3%, while office and administrative support declines 4.0%.
Limit. Projected gains and losses have multiple drivers. They cannot be attributed to AI alone or treated as guaranteed outcomes.
Study design, period & location in source
- Period covered
- 2025–2035
- Scope & method
- U.S. employment by industry and occupation.
- Where to look
- Occupational employment discussion and Table 1.
Artificial Intelligence exposure categories
U.S. Bureau of Labor Statistics
Finding. Categories describe relative exposure, not predicted replacement.
Limit. BLS explicitly excludes interpreting these as forecasts of employment, wages, adoption or productivity.
Study design, period & location in source
- Period covered
- 2026 release
- Scope & method
- U.S. occupations classified using several theoretical and observed-use measures.
- Where to look
- Definition, methodology and interpretation cautions.
Artificial Intelligence and the Future of Entry-Level Work
World Economic Forum & PwC
Finding. Proposes responses across job access, job design, talent pipelines and education.
Limit. A framework for action, not causal evidence of layoffs or proof that a particular intervention works.
Study design, period & location in source
- Period covered
- 2026 synthesis
- Scope & method
- Early-career pathways, education alignment and employer practices.
- Where to look
- Framework’s four dimensions and recommendations.
The Future of Jobs Report 2025
World Economic Forum
Finding. Employers anticipate demand for technology skills alongside cognitive, collaborative and care-related work.
Limit. Employer intentions are not realized outcomes; the sample does not represent every worker or business.
Study design, period & location in source
- Period covered
- Employer expectations through 2030
- Scope & method
- International employer expectations about jobs, skills and workforce strategies.
- Where to look
- Jobs and skills outlook; survey methodology.
World Robotics 2025: Top facts
International Federation of Robotics
Finding. 542,000 industrial robots installed in 2024; professional service-robot sales approached 200,000.
Limit. Units are not jobs displaced. Service figures are sample-based; industrial and service categories have different coverage.
Study design, period & location in source
- Period covered
- 2024 installations and sales
- Scope & method
- Industrial installations and service-robot supplier reporting; a robotics industry federation.
- Where to look
- Industrial and service robot top facts; service methodology notes.
World Robotics 2025: Service Robots executive summary
International Federation of Robotics
Finding. Logistics accounted for over half of professional units; 536 sales were recorded in the consumer care-at-home category.
Limit. Sample, not census. Consumer care-at-home excludes medical and other professional robots; sales are not installed stock.
Study design, period & location in source
- Period covered
- 2024 unit sales
- Scope & method
- Service-robot supplier sample, including professional and consumer categories.
- Where to look
- Professional service applications; consumer category AC3, care at home.
Task-Completion Time Horizons of Frontier AI Models
METR
Finding. Reports human-expert task duration at specified model success rates, including 50% and 80%.
Limit. Not autonomous runtime or whole-job competence. Long-duration estimates, especially beyond 16 hours, are uncertain.
Study design, period & location in source
- Period covered
- TH1.1 estimates; page snapshot reviewed September 2026
- Scope & method
- Primarily software, machine-learning and cybersecurity tasks with measurable completion criteria.
- Where to look
- Definitions, reliability selector and limitations below the chart.
The 2026 AI Index Report: Economy
Stanford Institute for Human-Centered AI
Finding. 88% of surveyed organizations reported AI use in 2025.
Limit. Not 88% of all businesses or workers. Adoption does not establish deployment depth, return on investment or job effects.
Study design, period & location in source
- Period covered
- 2025 adoption data
- Scope & method
- Organizational survey evidence assembled in the AI Index’s economy chapter.
- Where to look
- Economy chapter: organizational adoption.