All Sources

When we originally released GDPVal, which was just a few months ago, none of the models were yet on par with human experts. Months later, we have over an 80 percent win rate compared to human professionals.

Replicating the dominant rubric with three frontier models on identical tasks, we find a 3.6-fold divergence in mean exposure with agreement as low as 57%.

The research highlights a growing divergence between workers who can complement AI tools and those whose skills are substituted by them, amplifying within-occupation wage inequality.

A job is a bundle of tasks. The real question is not whether AI can perform one component of the bundle. It is whether that component can be separated from the rest at low cost... In 2013, a study by Carl Frey and Michael Osborne put the probability that accountants and auditors would be automated at 94 percent. A decade later, the US Bureau of Labor Statistics counts 1.6 million accountants and auditors employed, median pay of $81,680, and projects the occupation to grow another 5 percent through 2034... the argument that 'half of entry-level white-collar jobs be gone in five years' confuses task automation with the extinction of jobs.

Between 2011-2023, 18% of US workers were employed in jobs introduced since 1970. The wage premium is four times larger for new work associated with technological change than for other types of new work. The labor share has declined 10% in the US since the early 2000s.

Hundreds of millions of people use artificial intelligence (AI) for writing assistance.

Survey of 80,508 Claude.ai users (personal accounts). One fifth voiced concern about economic displacement. Perceived job threat correlated with observed exposure: for every 10pp increase in exposure, perceived threat increased by 1.3pp; top-quartile exposure workers mentioned worry 3x as often as bottom-quartile. Early-career respondents much more likely to express concern than senior workers. Mean productivity rating 5.1/7 ('substantially more productive'); 3% reported negative or neutral impacts, 42% no clear indication. 48% of users mentioning productivity cited scope (new tasks), 40% speed. Management occupations (mostly entrepreneurs) and computer/math groups showed largest gains; scientific and legal professions the mildest. 10% of respondents naming a beneficiary said employers/clients capture the surplus; only 60% of early-career workers said they personally benefited vs 80% of senior professionals. U-shaped relationship between reported speedup and perceived job threat: both those slowed and those sped up most are more anxious.

Three years after generative AI reached the market, there is no detectable employment effect for the most exposed occupations on UK data, regardless of which exposure metric is used. The estimates are noisy, the confidence intervals are wide, and neither measure produces a statistically distinguishable effect on aggregate employment.

U.S. labor productivity growth has accelerated, rising from 1.3% per year in the pre-pandemic expansion (2013–2019) to 2.2% in the post-pandemic period (2019–2025). Three technology-exposed groups — white-collar services, retail trade, and advanced manufacturing — are posting 3.2% to 3.9% annualized productivity growth. The rest of the private economy is at 0.1%. The pressure is likely to emerge first in entry-level white-collar work, where the same industries posting the strongest productivity gains are also showing weaker demand for junior labor.

Our proprietary survey of AI integration efforts finds that only 11% of management teams expect AI to replace existing software subscriptions. Far more expect the move to AI-centric architectures to reduce maintenance spending on legacy infrastructure, save on consulting fees, and increase the productivity of existing workers, reducing future headcount needs. Figure 4: Internal headcount cited as AI spending offset by 26% of respondents — the largest single category; IT services/consultants cited by 22%.

Workers initially employed in occupations that later decline by at least 25% demonstrate 0.4 lower future years of work, although this employment difference is mostly explained by other individual traits. These workers, conditional on controls, experience a 4.7% reduction in future cumulative earnings relative to starting earnings, akin to losing one year's worth of earnings over 2007–2024.

Metaculus community forecasts for 2030 and 2035: overall US employment -1.9%/-3.4%; most vulnerable AI-exposed occupations -11.4%/-17.2%; software developers -15.1%/-22.3%; financial specialists -8.1%/-15.3%; services sales -11%/-14%; lawyers -5.4%/-9.6%; designers -4%/-8.4%; K-12 teachers -1.3%/+1.3%; overall median wage -0.6%/+1.4%; percent workers using AI daily 52.5%/70.9%; new-grad unemployment ~10%/12%.

All 921 occupations (147.9M jobs) sort into four categories: 18% are at a higher short-term automation risk, 46% are less likely to experience near-term change, 12% could grow because of AI, and 24% may see declining employment as their task composition shifts but remaining jobs will still need workers. ChatGPT is used about 3x more in the kinds of jobs our framework identifies as most at risk of automation. Capability overhang by archetype: high-automation-risk jobs show 23.8% realized vs 90.0% theoretical exposure (66.2pp gap); jobs that grow with AI 22.7% vs 72.4% (49.7pp); reorganize 14.9% vs 67.1% (52.3pp); less immediate change 6.4% vs 27.4% (21.0pp). Since 2024Q1, unemployment rose most in jobs we predict to have less immediate change (+0.6pp) vs +0.3pp in higher-automation-risk, reorganize, and grow-with-AI groups — underscoring that exposure alone is a weak predictor of immediate labor market pressure.

What A.I. company executives are saying about their products — that they might lead to human extinction and almost certainly will lead to large-scale permanent disemployment — is so obviously 'bad messaging' that I would really urge people to consider that it's not a 'message' at all.

Regression-adjusted employment of early career workers in the most AI-exposed quintile of industry-state cells declined by 12% over the 10 quarters following the introduction of ChatGPT, even as employment in less-exposed industries has remained stable. I find that hires of these early career workers declined immediately by 9% in comparison with those in less exposed industries, and that they have not recovered over time.

The addition of the March 2026 CPS and the introduction of Anthropic's February usage metrics do not suggest any substantial changes. Occupational dissimilarity, industry dissimilarity, and our exposure and usage metrics all remain flat, lie within historical ranges, or continue along the trends they were already exhibiting. Currently, measures of exposure, automation, and augmentation show no sign of being related to changes in employment or unemployment.

Imas argues AI will trigger a post-commodity economy where spending shifts toward the relational sector (care, hospitality, craft, education) whose value is inseparable from human provenance. Evidence: Starbucks rolling back automation after it hurt satisfaction; experimental finding that human-made art gains 44% from exclusivity vs only 21% for AI-generated art.

Among currently employed respondents, 39 percent report that they are either using AI tools in their current job or have used AI tools in their jobs in the last twelve months. College graduates are more than twice as likely to have used AI tools at work in the past twelve months as those without a college degree (58.7 percent versus 22.9 percent). AI adoption rises from 15.9 percent among workers earning under $50,000 to 66.3 percent among those earning over $200,000 annually. Around 38 percent of employed respondents said that having training in how to use AI tools is important to them, yet only 15.9 percent report that their employer currently offers any AI training. Around 62 percent of all respondents believe the unemployment rate will increase over the next twelve months due to AI.

28% of employed U.S. adults use AI a few times a week or more; 13% use daily (up from 10% in 2024). 41% of employees report their organization has integrated AI tools. 23% in AI-adopting orgs report workforce reductions vs. 16% in non-adopting. Survey of 23,717 employed U.S. adults, margin of error +/-0.9pp.

Employment among software developers aged 22–25 has plummeted nearly 20% since 2024, even as their older colleagues' headcount grows. The pattern repeats in other jobs with higher levels of AI exposure, like customer service. Meanwhile, firm surveys indicate executives expect this trend to accelerate, with planned headcount reductions outpacing recent cuts. Generative AI reached 53% population adoption within three years, faster than the personal computer or the internet... the U.S. ranks 24th at 28.3%. Across multiple hospital systems, physicians reported up to 83% less time spent writing notes and significant reductions in burnout.

Long-lasting impacts: 10 years after a job loss, technology-displaced workers' real earnings were 10 percentage points below that of non-displaced workers. Short-run impacts: It can take one month longer for technology-displaced workers to find a new job; and their inflation-adjusted earnings take bigger hits (more than 3%) versus other workers (negligible effect). Recessions worsen outcomes: The effects of technology-related displacements are amplified (by three weeks of additional unemployment and a 5-percentage-point likelihood of subsequent joblessness). Goldman Sachs previously estimated that 6% to 7% of US workers (about 11 million people) could have their jobs displaced by AI.

I was somewhat surprised that the gap between sort of coding in general, which as we point out had something like 94% theoretical exposure, but then based on actual adoption, it was closer to 30% of the tasks across all the jobs in that pocket of the economy.

Here, through a series of randomized controlled trials on human-AI interactions (N = 1, 222), we provide causal evidence for two key consequences of AI assistance: reduced persistence and impairment of unassisted performance. Although AI assistance improves performance in the short-term, people perform significantly worse without AI and are more likely to give up.

Our analysis implies that AI substitution has reduced monthly payroll growth by roughly 25k and raised the unemployment rate by 0.16 percentage points over the past year, while augmentation has added about 9k to monthly payroll growth and lowered the unemployment rate by 0.06pp. This implies a net drag of 16k per month on payroll growth and a 0.1pp boost to the unemployment rate. These negative effects fall largely on less experienced workers, widening the entry-level-to-experienced wage gap by 1.3% and the unemployment rate gap by 0.6pp from their pre-pandemic averages.

AI will definitely eliminate some jobs, while it enhances others. Huge increase in AI-driven capital spending and construction by the five hyperscalers. In 2025, this number was $450 billion, and in 2026, it will be approximately $725 billion. There is a possibility that AI deployment will move faster than workforce adaptation to new job creation.

New research by Goldman Sachs economists finds that AI is already a measurable drag on the U.S. job market—erasing roughly 16,000 net jobs per month over the past year, with the pain falling hardest on Gen Z and entry-level workers. Goldman's breakdown shows AI substitution wiped out roughly 25,000 jobs per month in the past year, while augmentation added back about 9,000. The wage gap has similarly deteriorated, with Goldman's regression analysis estimating that a one standard-deviation increase in AI substitution exposure widens the entry-level-to-experienced wage gap by roughly 3.3 percentage points.

Total nonfarm payrolls rose by 571K in March 2026 to 157,775K.

Information sector employment at 2,772K, +3K month-over-month.

Unemployment rate 4.3% in March 2026, down 0.1 percentage points.

Most still do not see much evidence that A.I. is disrupting the job market. But they are starting to take seriously the possibility that it could someday soon. If it does, they are worried that policymakers are not ready to respond.

Autonomous coding agents are generating code at an unprecedented scale, with OpenAI Codex alone creating over 400,000 pull requests (PRs) in two months. CRA-only PRs achieve a 45.20% merge rate, 23.17 percentage points lower than human-only PRs (68.37%).

Adoption stood at about 18 percent of firms at the end of 2025. Prior to the question revision, the adoption rate had grown by 68 percent (3.9 percentage points) over the prior year but decelerated in Q2 2025. Over 20 percent of firms expect to use AI in the first half of 2026. The right panel of figure 2 shows that work-related GenAI adoption reported in the RPS stands at about 41 percent of the workforce, and non-work-related usage at about 50 percent of the population as of the latest survey in November 2025. The SBU estimates an employment-weighted firm AI adoption rate of around 78 percent and an LLM adoption rate of about 54 percent.

Some 15.6 million—or one-fifth of the nation's 70 million STARs—work in roles highly exposed to AI, meaning they reside in the top quartile of observed occupational task exposure among occupations.

U.S.-based employers announced 60,620 job cuts in March, according to Challenger, up 25% from 48,307 cuts announced in February. AI was the leading reason for cutting jobs, cited in 25% of announcements, followed by closings, restructuring and economic conditions.

NBER Working Paper 35046. Survey of 69 economists, 52 AI industry/policy professionals, 38 superforecasters, and 401 general public on AI's economic effects. Unconditional median economist forecasts: GDP growth 2.5% for 2025–2029, LFPR 61.0% for 2030 (vs 62.6% Jan 2025), 58.3% for 2050. Only 14.0% mean probability assigned to a 'rapid' AI progress scenario by 2030. Conditional on rapid scenario: GDP growth 3.5%, LFPR falling to 55.0% by 2050 (~10M lost jobs attributable to AI), wealth inequality reaching 80.0% held by top 10% by 2050, work hours AI-assisted rising from 3.35% (2024) to 10.1% (2030 unconditional) or 24.2% (2030 rapid). Variance decomposition finds expert disagreement driven primarily by different beliefs about economic effects of highly capable AI systems, not by disagreement about AI capability progress. Economists support retraining (71.8%) over job guarantees (13.7%) or UBI (37.4%); general public supports both.

GenAI reshapes wage inequality through a productivity-scarcity race: productivity gains widen wage disparities while reduced entry barriers compress them. Net effect is a modest increase in inequality.

Headcount reduction was the largest outcome in 45% of the deployments, but alternatives (hiring avoided, redeployment, no reduction) accounted for 55%. Agentic implementations showed 71% median productivity gains versus 40% for high-automation. 90-95% of food delivery customer service interactions fully automated by AI agent.

If progress continues, we can expect systems to be capable of carrying out projects that currently take people months. This shift will reshape how organizations run, how knowledge is created, and how people find meaning and opportunity.

Using realized worker-firm matches data for 138 million U.S. workers, Azar et al find the arrival of ChatGPT reduced average wages of employees at firms that can substitute AI for labor by 4.5% relative to firms that cannot.

In a rapid AI progress scenario, economists forecast a drop in labor force participation from its 2025 baseline of 62.6% to 59.1% in 2030 and 55% in 2050, with roughly half of this decline—equivalent to 10 million jobs—attributable to AI.

The activities that comprise transaction costs—learning prices, negotiating terms, writing contracts, and monitoring compliance—are precisely the types of tasks that AI agents can potentially perform at very low marginal cost. Once agents can indeed execute these functions effectively and cheaply, we will see significant shifts in the traditional make-or-buy boundaries that define firm organization and market structure.

The paper maps both the opportunities GenAI creates and the structural challenges it poses to workers across skill levels. It underscores the uneven distribution of risk, with routine cognitive roles facing the greatest exposure.

Despite faster growth, treated firms do not scale inputs proportionally. Their demand for external capital investment falls by 39.5% relative to the control group, while their demand for labor remains unchanged.

The paper emphasizes that white-collar and knowledge-work roles are now squarely in the displacement frame, departing from earlier automation waves that primarily affected manual labor.

92% of AI applications map to only 6.8% of 39,603 classified work activities. AI apps grew 6x from 2022-2024 but activity coverage expanded only 1.2x. 75% of AI market value concentrated in software/information tasks. 58% of AI apps target 'Create information' activities.

If AI dramatically boosts the productivity of high-paying knowledge work while leaving lower-paid service occupations largely unaffected, between-occupation inequality could widen even as within-task inequality narrows.

In 2022, there was essentially no relationship between a metro area's share of college graduates and where its unemployment rate stood relative to its own history. The correlation was -0.01. Three years later, that correlation is 0.26 — modest but meaningful across more than 300 metropolitan areas.

Industries with high volumes of structured, repetitive tasks face the sharpest near-term employment contractions.

Industry-wide vulnerability to job displacement is approximately 6%... 9.3 million jobs are vulnerable to job loss due to AI under our median adoption path, with a plausible range of 2.7 to 19.5 million

About 49% of jobs have seen at least a quarter of their tasks performed using Claude. High-tenure users have a 10% higher success rate. Early adopters with high-skill tasks have more successful interactions, identifying a channel through which skill-biased transformation may already be unfolding.

Two jobs with identical exposure scores can have completely opposite displacement risks depending on whether their tasks are complements, whether demand for their output is elastic or inelastic, and the incentives of the firm to invest in automation. The workers at greatest risk are not necessarily those with the highest average exposure, but those whose jobs are built around a small number of core tasks that AI can automate.

In weak-bundle occupations, AI automates some tasks and narrows the boundary of the job, leading to the standard task-substitution channel. In strong-bundle occupations where tasks are not independently reallocable, AI improves performance inside the job, but does not remove the human from the bundle.

After controlling for industry-level shocks we find that coder employment growth has been 3 percent lower since the introduction of ChatGPT. Cumulating over the roughly 3 years since November 2022 and using 5.735 million coder jobs as the base value, the implication is that roughly 500,000 additional coder jobs would have existed in the absence of large-scale LLM use.

The addition of the January and February 2026 CPS and the introduction of Anthropic's 'Observed Exposure' metrics do not suggest any substantial changes. Occupational dissimilarity, industry dissimilarity, and exposure and usage metrics all remain flat, lie within historical ranges, or continue along pre-existing trends.

6-7% of workers will be displaced during that transition period. In the US, AI can potentially automate tasks that account for 25% of all work hours. Goldman Sachs Research expects to see a 0.6 percentage point increase in the unemployment rate over a decade.

average individual wages are roughly unchanged for incumbents of occupations with no AI exposure, rise by around 4% at moderate exposure, and drop by as much as 35% for workers in the most exposed occupations

AI professionals earn $215,000, on average, in San Jose, CA — the highest in the country. But, with living costs 13% above average, that premium narrows. For comparison, in Dallas–Fort Worth, TX, $128,000 salaries stretch further with costs just 3% above average.

A lot of these claims are premature. Many of the people making them have an interest in selling a product. Right now I would describe the labor market as more of a wait-and-see environment. You cannot automate everything tomorrow because we simply don't have enough computing power to do it.

An event study documents an accelerating decline in employment of 22–25-year-olds in high-AI-exposure occupations, reaching 5.5 per cent by early 2025 relative to less exposed occupations within the same employers, while employment of workers over 50 rose by 1.3 per cent.

Two points of general agreement stand out: There's no measurable evidence so far that AI is putting Americans as a whole out of work, economists say. And while the victims of past workplace automation were mostly factory and trade workers, it's white collar jobs that are first in line for AI shake-ups today.

New business applications remain elevated but 'high propensity to hire' applications are in decline. SMBs are rapidly increasing tech spend (including AI) while payroll spend is flat or declining — consistent with AI-native solopreneurs substituting software for labor.

Survey of ~750 CFOs finds AI adoption widespread (58.5% in 2025, 85.4% expected 2026). Implied revenue-based labor productivity gains of 0.6% in 2025, 1.8% expected 2026, with largest gains in finance (>2%). Aggregate employment decline <0.4% in 2026 (~502K workers). Routine clerical roles declining; skilled-technical rising. Productivity paradox: reported gains 3x larger than implied revenue-based gains.

The evidence on how AI is affecting the labor market today is inconclusive, and claims about harmful impacts on particular groups of workers are premature. Initial evidence suggests that transitional disruption from AI to date is not outpacing recent technological changes.

I'm skeptical that simply slotting AI into human-shaped jobs will have the results people seem to expect... as long as we are in that regime, I expect disappointing productivity gains and relatively little real displacement.

89% of the respondents said they have no plans to cut staff or lay off workers because of AI. Just 3% said they have already cut staff, while 7% said they have plans to reduce their workforce within the next year.

AMIE's differential diagnosis included the final diagnosis in 90% of cases, with 75% top-3 accuracy. Blinded assessment suggested similar overall DDx and Mx plan quality between AMIE and PCPs. Human safety supervisors did not need to intervene to stop any consultations.

The U.S. Army's 18th Airborne Corps, using software from data company Palantir Technologies in a continuing string of exercises dubbed Scarlet Dragon, matched its own record from Iraq as the military's most efficient targeting operation ever. Thanks to AI, the corps achieved that with only 20 people, compared with more than 2,000 staffers employed in Iraq.

We find no systematic increase in unemployment for highly exposed workers since late 2022, though we find suggestive evidence that hiring of younger workers has slowed in exposed occupations

new grad hiring in large tech is down over 50% since 2019. For the first time in over decades, recent college grads have a higher unemployment rate than the national average. In addition, the 'underemployment rate' for recent graduates has risen to 42.5% (Q4 2025, New York Federal Reserve).

So far in 2026, AI has been cited for 12,304 job cut announcements, or 8% of job cut plans. In 2025, companies referenced AI for 54,836 announced layoff plans, 5% of total cuts during the year.

Are we finally seeing AI in the productivity data? A big upward revision to earlier data and strong Q4 bring us 2.2% above CBO's pre-pandemic forecast. Annual rates: 1 year: 2.8%, 2 years: 2.5%, 6 years: 2.2%

3% workforce reduction amidst record profits; investment banking revenue up 47%.

Companies with frequent AI use are about 4% more likely to take on additional staff. Two-thirds of the 3,500 firms that took part in the survey reported that their employees use AI.

In 2025, Ramp shipped 500+ features, reached $1B revenue, and did it all with 25 PMs.

According to the American Community Survey, 7.2 percent of early-career college graduates who majored in CS were unemployed in 2024—well above the 4.5 percent rate for all early-career graduates.

In spite of the current displacement narrative – job postings for software engineers are rising rapidly, up 11% YoY. There is little evidence of AI disruption in labor market data as of today.

The survey shows institutional adoption is accelerating, with 66% of respondents reporting their institution is currently leveraging AI, an increase from 49% year over year. Eighty-eight percent of respondents say they expect institutional AI use to increase over the next two years.

While we still do not find a meaningful relationship between productivity and AI adoption at the economywide level, companies that quantified productivity impacts of AI on specific tasks reported a median productivity gain of around 30%.

Project APE — Autonomous Policy Evaluation... has produced 204 papers — with 60 added in a single week... These are fully automated papers, like a version 1.0, with no human iteration whatsoever.

AI capex – roughly $650bn this year and potentially ~$2tn over three years – materially alters the equilibrium.

Today, AI apps have crossed 1 billion weekly active users and ChatGPT alone accounts for 900M of that.

Analysis of 4.5M+ non-prisoner federal civil cases (FY2005-FY2026) and 46M PACER docket entries. Pro se filing share rose from ~11% steady-state (FY2005-FY2022) to 16.8% in FY2025, with pro se case counts nearly doubling (23,210 pre-AI avg → 41,490 in FY2025). Rise concentrated in 'simple' NOS categories (civil rights, consumer credit, foreclosure); absent in patent/securities. Plaintiff-side only — defendant pro se counts fell. Pro se docket entries per court up 158% vs pre-AI mean by 2025Q2; per case up 38% (16.9 → 23.3). Represented entries per case also up 23% (18.2 → 22.5). Case durations and disposition mix unchanged. Pangram AI-text detector applied to 1,600 random complaints (200/yr, 2019-2026): AI-text share rose monotonically from 0.1% pre-AI (1 of 800 false-positive baseline) to 1.0% (2023), 3.5% (2024), 10.5% (2025), 18.0% (early 2026).

in 2026 we find that 43% of US workers use AI for their job compared with 32% among European workers in our surveys.

Across all benchmarks, examples collectively cover a limited 56.5% of the domain taxonomy but a substantially broader 85.4% of the skill taxonomy.

We first analyze 43 benchmarks and 72,342 tasks, measuring their alignment with human employment and capital allocation across all 1,016 real-world occupations in the U.S. labor market.

45% of leaders are willing to pay 11% to 15% more for strong AI skills

AI models successfully complete tasks that take humans approximately 3-4 hours with about a 50% success rate, increasing to about 65% by 2025-Q3. Most text-based tasks projected to reach 80-95% AI success by 2029. Based on 17,205 expert evaluations across 3,000+ O*NET tasks and 40+ LLMs.

Unemployment level 7,239K in March 2026.

As technology devalues labor, optimal policy shifts from steering innovation toward redistribution.

In the last two weeks, did this business use Artificial Intelligence (AI) in any of its business functions? Yes: 17.5%, No: 73.1%, Do not know: 9.4% (Cycle 202604, ref period Jan 26–Feb 8, 2026). First biweekly reading (Cycle 202524, Nov 2025): Yes: 17.3%.

AI's labor market impact still small — only 5-10K/month drag on net job growth. Only 2½% of jobs exposed to automation today. Baseline forecast: 6-7% of workers displaced (range 3-14%), lowering annual hiring by 1M jobs and raising unemployment ~½pp. US creates 30M gross new jobs/year; does not anticipate job apocalypse. Impact so far largely confined to tech sector but expects it to grow materially.

METR retracts original 19% slowdown finding, citing severe selection bias. Follow-up shows -18% speedup for original devs (CI: -38% to +9%), -4% for new recruits (CI: -15% to +9%). Authors state data gives 'unreliable signal' and are redesigning the study.

Employment in the computer systems design and related services sector has declined 5 percent. AI exposure associated with 0.28pp wage growth reduction for low-experience-premium jobs but 0.2pp increase for high-experience-premium occupations.

AI's impact depends on whether it automates or augments. For zero-experience-premium occupations, increased AI exposure is associated with -0.28pp reduction in wage growth. Wages in AI-exposed jobs not uniformly declining.

U.S. healthcare avoided $258B in admin costs. 50%+ of health plans and 25% of provider organizations use AI tools in admin workflows.

AI exposure metrics broadly agree with each other, but that they disagree with each other more on highly exposed occupations. The key point of disagreement between different AI exposure metrics is in the magnitude of exposure, not whether an occupation is exposed.

GPT-scored AI exposure for 342 BLS occupations on a 0-10 scale, validated against 6 academic metrics (Pearson 0.878 with Yale PCA, 0.885 with OpenAI). Employment-weighted industry exposure scores for 186 4-digit NAICS industries.

37% of tech professionals say their role has been 'redefined or restructured due to gen AI in the past two years.' 80% of respondents save at least one hour per day using generative AI. AI literacy job postings growing 70%+ year over year.

OpenAI average stock-based compensation reached $1.5M per employee in 2025. Research scientist total compensation ranges from $763K to $1.44M. The company raised $6.6B in October 2024 at a $157B valuation.

Full-year 2025 revenue of $430.9M, up 10.1% YoY. Active buyers down 13.6% in 2025. Spend per buyer reached $342, rising 13% YoY. Writing, translation, and simple programming categories declining ~20% due to AI substitution. High-value projects over $1,000 grew 23%. 2026 revenue guidance: $380M-$420M, a decline of 3-12% YoY.

Only 20% of customer service leaders have actually reduced agent staffing due to AI, yet 80%+ expect to reduce headcount over the next 18 months. Nearly 80% plan to transition agents into new roles. Survey of 321 CS leaders (Sep-Oct 2025).

AI adoption causally increases labour productivity levels by 4% on average in the EU, with no evidence of reduced employment in the short run. Wage gains may accrue disproportionately to highly skilled workers.

Customer service could see as much as 75% of interactions automated by 2026. AI-driven platforms could deliver primary investment advice to nearly 80% of retail investors by 2027. Over 85% of software developers now use AI coding assistants, delivering productivity gains of up to 60%.

69% of firms across US, UK, Germany, and Australia currently use some AI technology (US: 78%). LLM text generation most common (41%). Expected to rise to 75% in 3 years. Larger, more productive, higher-paying firms more likely to adopt.

Survey of ~6,000 executives across US, UK, Germany, Australia. 90%+ report no employment impact from AI over the past 3 years. Firms expect AI to reduce employment by 0.7% over the next 3 years (US: -1.2%). Employees, by contrast, expect +0.5% job creation.

Looking ahead three years, executives predict a 1.4 percent productivity boost and 0.8 percent output increase alongside a 0.7 percent employment reduction—while employees forecast a 0.5 percent employment increase.

68% (331 out of 485) of S&P 500 earnings calls in Q4 2025 cited AI — a new 10-year high. Financials and IT had highest counts (67 each). IT (94%), Financials (91%), Communication Services (89%) had highest sector percentages.

Revenue grew 2% year over year to $787.8 million for full year 2025. GSV per active client was $5,129 in Q4 2025, an increase of 7% year over year. AI-related skills grew by 109% year over year. Active clients were 785,000 as of December 31, 2025.

Job openings fell to their lowest level since December 2017. Professional & business services openings down 21.8% (-284,000). Financial activities openings down 25.1% (-86,000). Healthcare openings down 10.8% (-152,000) since October.

Survey respondents said that AI adoption had led to elimination of 11% of jobs and an additional 12% were left unfilled. This is partially offset by 18% new hires, resulting in a 4% net job loss globally.

Cutting ~20,000 jobs (8% of workforce) by 2026 in restructuring tied to automation; broader plans for 60,000 total reductions including Banamex IPO.

35.9% of US workers used generative AI by December 2025; adoption is accelerating but remains unevenly distributed across industries and skill levels.

No measurable change in job openings or aggregate employment in AI-exposed occupations; adjustment happening through task reallocation, not broad displacement.

Pressure concentrated in entry-level segments of highly exposed occupations; adjustment happening at the margin through task reallocation.

Small but statistically significant positive wage effects found across AI-exposed occupations; 35.9% of US workers used generative AI by December 2025.

a 10 percent increase in the minimum wage increases robot adoption by roughly 8 percent relative to the mean

In a randomized experiment with 1,174 adults ages 25-45, AI access closed approximately three-quarters of the education-based productivity gap: higher-education participants outperform lower-education participants by 0.548 standard deviations without AI; with AI, this gap falls to 0.139 standard deviations.

The United States has seen 50% growth in AI-related postings over the past year, with AI now appearing in 2.5% of all job postings.

Without AI, higher-education participants outperformed lower-education participants by 0.548 standard deviations. With AI access, this gap fell to 0.139 standard deviations—closing about 75 percent of the baseline productivity difference.

Synthesizes firm-level survey data from the US, UK, Germany, and Australia measuring AI's realized and expected impacts on sales per employee and employment. Data collected Feb-Apr 2025 and Nov 2025-Jan 2026.

Job openings at 6,882K in February 2026, gradual decline from 2024 peaks.

AI use impairs conceptual understanding, code reading, and debugging abilities. The AI group averaged 50% on the quiz, compared to 67% in the hand-coding group.

36.8% of individuals across OECD used generative AI in 2025. 20.2% of firms used AI in 2025 (up from 8.7% in 2023). 41.1% of employed individuals used generative AI. 52.0% of large firms vs 17.4% small.

I've been warning about this risk for over a year; now, the people building the technology are putting timelines on it.

6.1 million U.S. workers face both high AI exposure and low adaptive capacity. Medical secretaries and administrative assistants (831,000 workers) stand out as one of the largest occupations in this high-risk category. About 86% of these workers are women.

Administration comprises 25% of all healthcare costs, making it a key target for automation. Organizations are using AI for documentation, coding, scheduling. Ambient documentation frees up to four hours of clinician time.

Only about 43% of US workers reported regularly using AI at work last year

AI is redistributing economic value away from routine cognitive tasks toward judgment, coordination, and specialized knowledge.

In het derde kwartaal van 2025 werkten ruim dertien procent minder jongeren in de meest vatbare beroepsgroepen dan in het vierde kwartaal van 2022, terwijl de werkgelegenheid in andere beroepsgroepen juist drie procent steeg.

6.1 million workers in AI-exposed roles lack adaptive capacity; 86% are women. Vulnerability concentrated in college towns and state capitals in Mountain West and Midwest.

Of the 37.1 million workers in the top quartile of AI exposure, 26.5 million are in occupations that also have above-median adaptive capacity. 6.1 million workers (4.2%) face high exposure + low adaptive capacity; 86% are women.

raising average wages by 21 percent

A one-standard-deviation increase in Generative AI exposure is associated with an 8% decline in job postings and a 0.6% decline in the hourly wage rate.

AI skills significantly increase interview invitation probabilities by approximately 8 to 15 percentage points, compared with candidates without such skills

Computer/mathematical tasks account for ~33% of all Claude.ai conversations and ~50% of API traffic, indicating concentrated AI impact on tech-adjacent roles.

49% of jobs now use AI for at least 25% of their tasks (up from 36% in early 2025); widespread adoption could boost US labor productivity by 1.0-1.8 percentage points annually.

49% of jobs now use AI for at least 25% of their tasks (up from 36% in early 2025); augmentation (52%) has overtaken automation (45%) as primary use pattern.

Education task exposure jumped from 11% to 49% — a 4.5x increase driven by multimodal AI, advanced reasoning, and agentic systems. Based on reassessment of 18,000 tasks across 1,000 O*NET occupations.

Unemployment risk in AI-exposed occupations rose beginning early 2022; graduates with AI-exposed curricula have higher first-job pay and shorter job searches post-ChatGPT.

Unemployment risk in AI-exposed occupations rose beginning early 2022; graduates with AI-exposed curricula have higher first-job pay and shorter job searches post-ChatGPT.

95% of generative AI projects have failed to deliver measurable returns (citing MIT). PwC: 56% of CEOs said AI failed to boost revenue or lower costs. Layoffs are anticipatory, not performance-driven.

Around 40% of global employment is exposed to AI; in advanced economies 60%. Demand for new skills is rising with 10% of jobs in advanced economies requiring at least one new skill.

Emerging evidence that generative AI adoption is reducing entry-level hiring, especially where tasks are automatable rather than complementary to humans.

Employment 3.6% lower in AI-vulnerable occupations in regions with high AI-skill demand after 5 years; geographic divergence driven by skill concentration.

Job postings with 4+ new skills pay up to 15% more (UK) and 8.5% more (US); middle-skill roles being squeezed while high-skill workers gain.

95% of customer interactions projected to be AI-powered by 2028. Current AI resolution rates at leading companies exceed 55% without human escalation.

By 2040, output is only 4% higher than it would have been without the growth acceleration, and by 2060 the gain is still only 19%. A key reason for the slow acceleration is the prominence of 'weak links' (an elasticity of substitution among tasks less than one).

In the AI group, screen readings fell from 109,692 to 61,248—a 44 percent workload reduction—while screen-detected cancers rose from 262 to 338, a 29 percent increase in detection.

New book releases tripled over this period, but average quality declined... the authors estimate a potential steady-state consumer surplus gain of 25–50 percent from market expansion.

Retail trade projected to decline 1.2% as automation and e-commerce reduce sales occupations. Mining/extraction declining 1.6% driven by robotics and drones. Total employment projected to grow 3.1% (2024-34), much slower than prior decade.

Trinity Health is laying off 10.5% of its revenue cycle management staff across 15 states and 92 hospitals. Non-patient-facing revenue cycle functions will be transitioned to an external partner.

Nearly 3 in 4 (74%) companies plan to deploy agentic AI within two years. 34% are already starting to use AI to deeply transform their businesses. 84% of companies have not redesigned jobs around AI capabilities.

71% of organizations regularly use generative AI in at least one business function (up from 65% in early 2024). 92% plan to invest in generative AI over next three years. 66% report productivity gains.

The high demand for new IT skills may not necessarily translate into a one-for-one rise in demand for IT and AI specialists, especially as many IT tasks may progressively be automated by AI.

There is no evidence that job postings for junior roles within occupations most exposed to AI have declined more than postings for senior positions. Postings for both levels of seniority have been falling roughly in parallel since their peak in Spring 2022, with the decline in junior positions stabilizing faster.

Nearly 40% of AI time savings are lost to rework, including correcting errors, rewriting content, and verifying outputs from one-size-fits-all AI tools. Only 14% of employees consistently get clear, positive net outcomes from AI.

CEO Brian Moynihan: 'We have 18,000 people on the company's payroll who code, and we've — using the AI techniques, we've taken 30% out of the coding part of the stream of introducing a new product or service or change that saved us about 2,000 people.'

94% of higher education workers use AI tools for work, but only 54% are aware of institutional AI policies. 67% identified six or more urgent AI-related risks, including job loss. Survey of 1,960 staff and faculty across 1,800+ institutions.

19% of respondents said they are actively pursuing different careers (7%) or considering changing plans in the near future (12%) due to AI-driven transformation.

BLS projects a further 6% decline in programmer roles through 2034. Computer programmer employment (routine coding roles) fell ~27.5% in roughly two years following ChatGPT's release — one of the largest two-year drops in any occupation tracked by BLS. Software developer employment remained flat.

Employers cited AI as a direct cause for 55,000 job cuts in 2025, concentrated in customer service, data entry, and back-office operations.

automation can depress demand and push AI-driven growth toward the lower end of current forecasts. [...] the conditions needed for growth to actually turn negative are likely too unrealistic to hold in practice

Employment share of young workers (20-24) in most AI-exposed occupations fell from 16.4% to 15.5%; decline driven by fewer people transitioning into employment, not layoffs.

If fully translated to unemployment, AI-driven employment declines among young workers would explain only 0.1 percentage point rise in aggregate unemployment.

Employment share of young workers (20-24) in most AI-exposed occupations fell from 16.4% to 15.5% between November 2022 and September 2025.

Employment share for AI-exposed occupations fell from 16.4% (Nov 2022) to 15.5% (Sep 2025). Workers age 22-25 in most AI-exposed occupations experienced 13% employment decline since 2022, driven by fewer workforce entrants rather than layoffs.

AI/ML engineers median total compensation reached $185K in 2025, with top-tier researchers commanding $500K-$1M+. The premium over general software roles widened to 35%.

Widely-used exposure indices, which aggregate task-level automation risk using linear formulas, will overstate displacement when tasks are complements. Labour income can rise under partial automation because automation scales the value of remaining bottleneck tasks.

35.9% of US workers used generative AI by December 2025; small positive wage effects, no significant decline in job openings.

AI adoption led to moderate productivity increases — an 8.5 percent increase in coding activity and 8.7 percent faster task completion — with no measurable quality declines. These productivity gains did not translate into increased output, changes in task composition, or effects on employment.

Perceived 'identity fit' — whether a new skill feels compatible with a person's sense of self — dominated re-skilling decisions, often outweighing beliefs about wages or employer demand.

Freelance marketplace spend collapsed from 0.66% to 0.14% of GDP after ChatGPT launch.

Pixiv illustrator uploads down 10.1% post-AI; commercial creators saw -14.3% decline.

EU firms see +4% productivity from AI adoption; no employment decline detected.

Gender gap in AI adoption: women adopt less; 5.3 percentage point gap in UK data.

Combined headcount fell to 1.09 million — lowest since 2021, down ~10,600 from prior year, largest annual reduction since 2016.

AI adoption for production rose from 4.6% (early 2024) to 10% (Sep 2025). Broader measure (any business function): 17.3% (Nov 2025). 78.4% annualized growth rate. 80.9% of non-adopters cite inapplicability.

falling labor share accounted for 16% of U.S. real wage growth between 1954 and 2019

A Deloitte survey reveals how AI and human collaboration can help close talent gaps, speed upskilling, and transfer knowledge as demographic changes reshape the workforce.

Each year of AI model progress reduced task time by 8%. Access to any AI model increased Earnings Per Minute by 81.3%. Continued scaling could boost US productivity by ~20% over the next decade.

Challenger, Gray & Christmas reports AI has been cited as the reason for 54,694 planned layoffs in 2025. Overall job cuts topped 1 million in 2025.

No economy-wide employment disruption detected since ChatGPT's release, but this masks occupation-specific impacts for early-career cognitive workers.

AI was the most-cited topic in S&P 500 earnings calls in 2025. Companies mentioning AI in workforce context outperformed peers by 8.2 percentage points on average.

46% of US workers reported using AI tools at work in Q4 2025, up from 33% in Q1 2025. Adoption driven by white-collar professional and technical roles.

AI chatbots had no statistically significant impact on hours worked or wages earned for lawyers. Law graduate employment in 2024 reached 93.4% — the highest rate on record.

Despite early signs that AI is beginning to affect entry-level workers, labor statistics have yet to show lawyers being displaced. Law graduate employment in 2024 reached 93.4% — the highest rate on record.

Mountain region leads US AI adoption at 42%, followed by Pacific at 39%. Southeast lags at 22%. Regional adoption gaps correlate with widening wage divergence.

AI adoption in core business functions ranges from 1.9% in Japan to 6.1% in the United States in 2024. Among SMEs using generative AI, only 29% report using it in their core activities. Annual labour productivity growth stemming from AI could range from 0.4 to 1.3 percentage points for the US over the next decade. Around a quarter of OECD workers are exposed to generative AI. 50% of SMEs report employees lack the skills to use generative AI.

Since January 2023, salaries at the top quartile of the wage distribution have grown by more than 30%. Over the same period, salaries at the bottom quartile have risen by only around 10%.

54,694 job cuts cited AI as a reason in 2025, concentrated in tech, finance, and customer service sectors.

78% of organizations report using AI in at least one business function. Enterprise AI spending grew 45% YoY globally.

Education support roles face moderate displacement risk while core teaching remains protected. Administrative and content creation roles most affected by AI adoption.

40% of employers globally expect to reduce headcount as AI automates tasks. Net displacement estimates range from 5-14% of current roles by 2030.

200,000+ European banking jobs could vanish by 2030 — roughly 10% of workforce at 35 major European lenders (2.12M staff). Efficiency gains of 30% quoted by banks.

61% of S&P 500 companies cited AI in Q4 2025 earnings calls, maintaining near-record levels from Q3.

306 S&P 500 earnings calls cited AI in Q3 2025 (then-record), significantly above the 5-year average of 136 and 10-year average of 86. IT (95%) and Communication Services (95%) had highest sector percentages.

AI can already replace 11.7% of the U.S. workforce, worth $1.2 trillion in wages. Exposure is distributed across all 50 states, not confined to coastal tech hubs. South Dakota, NC, Utah show higher values than California.

57% of US work hours are technically automatable. Physical tasks comprise 50%+ of hours for 40% of the US workforce. Robot unit costs ($150-500K) must fall to $20-50K for mass physical automation adoption. At least 14% of employees globally may need career changes by 2030.

Currently demonstrated technologies could automate activities accounting for about 57 percent of US work hours today. AI agents could perform tasks occupying 44 percent of US work hours, while robots could handle another 13 percent. Roles with the highest potential for automation make up about 40 percent of total jobs.

By 2026, 20% of organizations will use AI to flatten their organizational structure, eliminating more than half of current middle management positions in those organizations.

75% of knowledge workers now use AI tools. Adoption in enterprises with 500+ employees reached 60%, up from 40% a year prior.

Industries more compatible with AI experienced an increase of over 10% in firm formation following ChatGPT's release. New firms are more likely to survive, grow faster, and attract more educated workers.

Median annual wage for all occupations: $49,500. AI-exposed administrative and clerical wages grew 1.1% vs. 3.8% for non-AI-exposed occupations.

Record 306 S&P 500 companies (61%) cited AI in Q3 2025 earnings calls — a new 10-year high, surpassing the prior record of 292. 95% of IT and Communications sectors mentioned AI. AI-citing companies outperformed 13.9% vs. 5.7%.

LLMs disrupted labor market signaling on Freelancer.com coding jobs (2.7M applications, 61K posts). Employer WTP for 1 SD signal increase fell from $25.67 to $14.85 (42% decline). Counterfactual no-signaling equilibrium: 5% avg wage decline, 1.5% hiring reduction, 4% worker surplus loss, top-quintile hired 19% less, bottom-quintile hired 14% more. 14% of post-LLM apps used AI writing tool.

Early-career workers (ages 22-25) in AI-exposed occupations experienced 16% relative employment declines, controlling for firm-level shocks. By September 2025, employment for software developers aged 22-25 declined nearly 20% compared to its peak in late 2022.

We find little difference in annual salary trends by age or exposure quintile, suggesting possible wage stickiness. Labor market adjustments are visible in employment more than in compensation.

Workers aged 22 to 25 have experienced a 6% decline in employment from late 2022 to September 2025 in the most AI-exposed occupations, compared to a 6-9% increase for older workers.

Overall employment continues to grow robustly, but employment growth for young workers has been stagnant since late 2022. Declining employment in AI-exposed jobs drives stagnant overall employment growth for 22- to 25-year-olds.

UnitedHealth deployed over 1,000 AI applications across insurance, care delivery, and pharmacy. AI chatbots handled over 65 million customer calls in 2024. Invested $1.5B in AI targeting ~$1B in annual operating cost savings.

Nearly half of all respondents (49%) use AI daily for client work. 69% don't feel fully prepared for an AI-driven creative industry. 1 in 3 expect design roles to transform this year. Survey of 1,780 global creatives.

Software output increased by 39% after the AI agent became the platform's default code generation mode. Quality did not deteriorate. Senior developers benefit most from agents; junior developers benefit most from autocomplete.

Claude speeds up individual tasks by about 80%. Extrapolating suggests 1.8% annual US productivity growth over the next decade.

32% of companies expect AI to reduce workforce by 3%+ within next year. Median 17% of respondents report workforce declines within functions from AI use. 88% of organizations now use AI in at least one business function.

Computer graphic artist postings down 33% in 2025, photographers and writers down 28%, journalists down 22%, against an 8% overall market decline. Analysis of 180M global job postings from company websites using 650 normalized job titles.

the median expert predicted 2% growth in white-collar jobs between January 2025 and December 2030, compared to 6.8% pre-existing trend; one in four experts foreseeing net job losses

Quarterly dashboard tracking generative AI adoption among U.S. working-age adults. Feb & May 2025 values estimated from tracker chart; all other values from published working paper.

Study of 11 occupations in Denmark: essentially zero effects on earnings and hours worked; confidence intervals rule out effects larger than 2%. AI adoption linked to ~4% occupational switching.

88% of organizations use AI in ≥1 function (up from 78%). Gen AI adoption at 71%. But only 7% fully scaled, and just 6% report EBIT impact ≥5%. 62% experimenting with AI agents.

Survey of 2,500 C-suite executives finds 90% report no significant net employment impact from AI adoption as of Q3 2025, though task reallocation is widespread.

Job postings for high-AI-substitution occupations fell 12% relative to low-substitution roles post-ChatGPT; effect grew from 6% in year one to 18% by year three. Based on 285 million Lightcast job postings.

Plans to reduce approximately 20% of full-time workforce by 2028, citing AI capabilities in operations and compliance.

23% of employed workers used generative AI at work weekly as of late 2024. Share of work hours with GenAI increased from 4.1% (Nov 2024) to 5.7% (Aug 2025). GenAI may have increased aggregate labor productivity by up to 1.3%.

Health insurance companies deploying AI to combat rising costs driven by hospital use of AI for billing and coding.

AI agents perform near the floor on RLI, with the highest-performing agent achieving an automation rate of 2.5%. 240 real freelance projects across game dev, design, architecture, data analysis, and animation.

Financial institutions could see 14% headcount reduction over next 3 years (worst of any category surveyed); only 11% of clients actively cutting employees due to AI currently.

Amazon plans to cut 14K corporate jobs (10% of white-collar workforce). UPS reduced management by 14K. Target cut 1,800 corporate roles. Nearly 2M Americans unemployed 27+ weeks.

Whether AI destroys jobs or transforms them, young people still need fundamental supports: skill-building, career navigation, and mentorship.

Less than 1% of announced layoffs in the first half of 2025 are attributable to productivity gains from AI; most (79%) have been unrelated to AI.

Gartner projects 20-30% of customer service agents will be replaced by generative AI by end of 2026. AI adoption in CS jumped from 46% to 61% in one year.

Workers with AI skills now command a 56% wage premium over non-AI peers, up from 25% just one year prior. AI roles median compensation reaches $157K.

Analysis finds no significant macro-level job displacement from AI as of late 2025, though structural shifts are emerging in specific sectors.

Operations and support staff to fall by at least 10% over next five years despite growing business volumes.

AI-skilled workers command a 56% wage premium, up from 25% in prior year. AI roles median compensation at $157K vs. $49,500 all-occupation median.

Large-scale field experiments find treatment effects 0-16.3% on sales, mapping directly into TFP improvements.

Indeed analysis finds 26% of jobs face radical transformation from AI. Paralegals face 80% automation risk by 2026. Legal research and contract review most affected.

The overall effect of industrial robots on wages is close to zero and insignificant. 53 papers, 2,143 estimations.

Survey of 180 Fortune 100 executives and 12,000 knowledge workers. 96% of organizations not seeing dramatic improvements in efficiency, innovation, or work quality from AI. Costing Fortune 500 ~$98B/year in lost returns. Workers report feeling 33% more productive individually, but organizational metrics flat.

Nationally representative survey of 25,000+ working-age adults finds generative AI adoption reached 55.9% overall and 40.7% at work by November 2025, outpacing PC and internet adoption at comparable points post-launch.

BofA's Erica AI assistant surpassed 2 billion client interactions with a 98% resolution rate. Now handles 56% of all customer inquiries without human escalation.

78,000 tech sector job losses attributed to AI in the first half of 2025, with QA testing, junior development, and IT support roles most affected.

AI chatbots for customer service increased sales by 16%; AI-generated product descriptions increased sales by 2.05% in field experiments.

No economy-wide employment disruption tied to AI since ChatGPT's debut. Measures of AI exposure show no clear relationship with changes in employment or unemployment.

Despite fears of imminent AI jobs apocalypse, the overall labor market shows more continuity than disruption since ChatGPT's launch.

At least 50% of tasks are automated in only 7.3% of education and library occupations — the lowest of any occupational group. Survey of 20,262 U.S. workers fielded March-April 2025.

542,000 industrial robots installed in 2024 — more than double the number 10 years ago. 4,664,000 robots operational worldwide. US installations at 34,200 units. Global market value $16.5 billion.

15% of US jobs — about 23 million total — are at high risk of displacement due to automation, concentrated in manufacturing, transportation, and food preparation.

US manufacturing lost 78,000 jobs over the past year with 12,000 cuts in August alone. Manufacturing employment at lowest level since onset of COVID-19 pandemic.

Among organizations with extensive agentic AI adoption, 45% expect reductions in middle management layers.

About half of respondent firms in the Philadelphia Fed's Third District are using generative AI. Survey conducted late October 2025.

In sharp contrast to the past two centuries, AI will shift relative demand toward occupations with lower education, lower wages, and a greater share of male workers.

39% of US adults used AI by August 2024. Small-firm adoption pace continues rising. Firms with 250+ employees at 11% usage; smallest firms (1-4 employees) at 7%.

Survey of 1,150 US workers: 40% received AI-generated 'workslop' in the past month; each instance cost ~2 hours to deal with. 18% of AI users admitted sending low-quality AI output. Estimated cost: $186/worker/month or ~$9M/year for a 10,000-person org.

Freelance marketplace spend fell from 0.66% to 0.14% of total business spend (Q4 2021 to Q3 2025); substitution ratio: $1 reduced freelance spend → $0.03 in AI spend.

Census Bureau rigorous measure finds 9.7% of firms actively using AI, while broader surveys report 78%. Adoption gap driven by definitional differences in what counts as AI use.

Only ~25% of Information sector businesses using AI; ~2.5% in Accommodation/Food — adoption patterns driven more by workforce composition than income.

Productivity gains 20-60% in controlled RCTs, 15-30% in field experiments; AI exposure measures converge toward high-wage jobs being most exposed.

18% of CFOs have already eliminated roles due to AI. Of those, 88% of cuts were in accounting, 38% in FP&A, 33% in treasury. Finance teams are the earliest white-collar AI displacement.

Revere Health eliminated 177 positions (billing, coding, accounts receivable) after partnering with IKS Health for AI-driven claims processing and automated billing. Largest documented healthcare admin AI layoff.

Peer-reviewed study finds AI can now automate substantial portions of creative work. Freelance writing gigs dropped 30% post-ChatGPT launch.

By July 2025, 18 billion messages were being sent each week by 700 million users. 81% of work-related messages involve obtaining information or making decisions, not task automation.

AI usage heavily concentrated by income and geography globally; within US, workforce composition matters more than income for adoption patterns.

In about 40% of employment in exposed occupations, at least 50% of tasks will be replaceable. Average labor-cost savings ~25% from current tools, potentially 40% as systems improve.

Roughly 10% of US work affected in the short run; AI will increase productivity and GDP by 1.5% by 2035. Jobs that AI can completely replace saw 0.75% employment fall 2021-2024.

50% are with agents, 50% are with humans. From 9,000 heads to about 5,000 because I need less heads. Support costs reduced by 17%.

In jobs with high AI exposure, employment for 22- to 25-year-olds fell 6% between late 2022 and July 2025. Software developers saw a 20% early-career decline. Employment among workers 30 and older grew 6-13%.

Software developer employment projected to grow 17.9% (2023-2033). Despite AI exposure, this occupation is unlikely to experience employment decline because robust software needs are expected to support continued demand.

IT unemployment rate rose from 3.9% to 5.7% between late 2024 and mid-2025, the highest level since 2021, driven by AI-related restructuring.

67% of providers believe AI can improve claims processing, but only 14% have implemented AI tools. Among adopters, 69% report reduced denials. Survey of 250 providers.

Analysis of 62M U.S. resumes: junior employment declined sharply at AI-adopting firms while senior employment was largely unchanged — described as 'seniority-biased technological change.'

Sufficiently advanced AI could dramatically raise growth rates and lower labor share, breaking Kaldor Facts. Wages may rise or fall depending on returns to scale and direction of technical change.

37% of software development tasks show AI involvement. Writing code is 18% of actual AI usage in software development; most use is for understanding, debugging, and reviewing code.

The framework yields several predictions: larger average firm size, greater industry concentration, and reduced local managerial autonomy. Transformative AI sharply expands what counts as codifiable local knowledge. In the absence of active countermeasures, transformative AI may lead to significantly more centralization of decision-making.

Global banks expected to shed 200,000 roles over 3-5 years (~3% headcount drop). Citigroup: 54% of banking jobs have high automation potential. AI could lift bank pre-tax profits 12-17% by 2027.

Audit of 500 AI vendors finds surveillance pay spreading across healthcare, customer service, and logistics.

37.4% of US workers used AI on the job in the past week. Adoption highest among college-educated (52%) and workers earning $100K+ (58%). Survey of 25,000 workers.

Companies hiring AI talent face a 30% salary premium as demand far outstrips supply, with experts warning that the cost of delayed AI hiring will only increase over time.

Unemployment rate for 20- to 30-year-olds in tech rose nearly 3% since early 2024. Entry-level job postings in the US diminished about 35% since January 2023. Goldman Sachs predicts AI will displace about 6-7% of the total workforce.

70% of providers and 80% of payers now have an AI strategy in place or in development. RCM is the top AI use case, with ambient documentation at ~20% full rollout and ~40% in pilot. Nearly half of provider executives said revenue cycle management was a top three IT investment priority. Survey of 228 US healthcare provider and payer executives.

~11% decline for early-career customer service workers; declines concentrated in roles where AI automates rather than augments tasks.

Young tech workers (20-30) unemployment risen ~3pp since start of 2025; tech employment share fallen below pre-pandemic trend.

Physically demanding or outdoor work unlikely to be affected by current AI. Projects 50,000-100,000 humanoid robot shipments in 2026. Unit economics expected to reach $15,000-$20,000 per robot.

Occupations with higher AI exposure experienced larger unemployment rate increases between 2022 and 2025 (r=0.47). Computer/math occupations (exposure ~80%) saw steepest rises. GenAI-intensive occupations: r=0.57.

Overall tech postings 36% below pre-pandemic baseline. Software engineer postings down 49%. Of 149 tech titles, only 19% exceeded pre-pandemic levels.

52% of the Class of 2023 was underemployed a year after graduating. Entry-level postings in high-AI-exposure occupations fell from 2022 to 2025, while postings requiring 6+ years of experience held steady or rose.

Job postings for non-tech roles that require AI skills offer 28% higher salaries — an average of nearly $18,000 more per year. The divide between AI-skilled and non-AI-skilled workers is widening.

Analysis of 1.3B job postings finds AI skills command 28% salary premium (~$18K/year); 43% for workers with 2+ AI skills. 51% of AI-skill postings now outside IT/CS with 800% growth in non-tech gen AI roles since 2022.

Teacher aide and educational support employment projected to decline 3%. Post-secondary teacher demand flat. Private tutoring category declining.

Computer graphic artist job postings declined 33% from pre-ChatGPT levels. Illustration and copywriting postings down 40%+. AI-creative hybrid roles growing.

Amazon deployed its 1 millionth robot across over 300 fulfillment centers worldwide. CEO Andy Jassy acknowledged that AI will result in fewer people doing some automated jobs, while new roles in robotics maintenance grew 30% at next-generation sites.

Real median weekly earnings for workers aged 16-24 in professional services fell 5.2% from 2023 to 2025.

Medical records, billing, and coding specialist employment projected to decline 8% through 2033, the steepest decline among healthcare occupations.

US firm AI usage in production reached 10.2%, with 14.8% in testing. Large firms (500+ employees) at 28%. Adoption accelerating across all sectors.

AI jobs pay more than similar jobs that do not focus on AI, with a typical premium of 25%. The share of AI jobs among new job listings increased 123% from 2023 to 2024.

Significant decline in job availability and earnings for freelancers in AI-affected creative occupations; some workers adapted by transitioning to programming.

High-skill freelancers disproportionately affected by AI — top performers were impacted most, contradicting conventional wisdom that expertise insulates workers.

Professional services occupations face the largest task-level AI exposure among high-income economies, with 18-25% of tasks fully automatable by current LLMs.

Solo-founder share of new startups rose from 23.7% (2019) to 36.3% (H1 2025). Acceleration from 30.5% to 36.3% in 2024-2025 coincides with AI coding assistants going mainstream.

Computer and mathematical occupations in San Jose-SF-Oakland MSA: $178K median. National median: $108K. Gap up from 48% to 65% since 2022.

automation has raised wages and reduced employment in occupations where it eliminated inexpert tasks, but lowered wages and increased employment in occupations where it eliminated expert tasks. These effects are distinct from—and in the case of employment, opposite to—the effects of changing task quantities.

Job growth remained robust even in AI-exposed occupations (38%). Degree requirements fell 7pp for AI-augmented jobs. AI-exposed professions saw 55% faster skill churn in job postings.

27.1% of companies expect artificial intelligence to lead to job cuts in the next five years. Only 5.2% of companies anticipate additional jobs, while two-thirds expect no change.

Calibrated to U.S. data, the model predicts more than threefold improvements in productivity in the some-AI steady state, alongside a long-run employment loss of 23%, with half this loss occurring over the initial five-year transition.

Graphic designer employment projected to decline 4% through 2033. Technical writers projected flat. New AI-creative hybrid roles growing.

26% increase in completed pull requests, 13.55% increase in commits, 38.38% increase in builds — but build success rate fell by 5.53 percentage points.

Using O*NET and LightCast job postings data (2019-June 2024), the study found a 24% decrease in generative AI-exposed skills per firm per quarter for automation-prone occupations, while augmentation-prone occupations saw a 15% increase. Demonstrates the dual displacement/complementarity impact of generative AI.

Less than 10% of firms using AI regularly as of mid-2025 (BTOS data); ~20% adoption in professional, scientific, and technical industries.

Tech/info/media employment stopped growing at end of 2022, coinciding with ChatGPT release; less than 10% of firms using AI regularly as of mid-2025.

Non-routine cognitive job workers now face rising unemployment risk, surpassing non-routine manual workers for the first time.

Klarna reversed its AI-only CS strategy after customer satisfaction dropped 22%. CEO Siemiatkowski: 'we went too far.' Now operates hybrid model with AI handling ~66% but humans for complex cases.

Survey of 335 freelance creative workers: 73% believe GenAI is altering quality of work; 68% feel job security reduced; 61% say value placed on their work has decreased; 55% report decline in financial compensation.

678 occupations having on average 23% of their tasks exposed to Generative AI

Human-AI teams show roughly 73 percent higher productivity per worker for ad copy, though human-human teams remained superior for images.

Productivity growth in AI-exposed industries including financial services nearly quadrupled from 7% to 27% (2018-2024); AI-exposed sectors see 3x higher revenue-per-employee growth.

AI-assisted coding jumped from around 5% in 2022 to nearly 30% in the last quarter of 2024. Quarterly output has increased by 3.6%. Experienced programmers benefit more; beginners hardly benefit at all.

GitHub Copilot saves 26 minutes per day across 20,000 UK civil servants.

Exposed startups increased productivity, scaled faster, accelerated through financing rounds. VC shifted toward frequent, smaller investments, boosting new firm formation.

GenAI boosts innovation by enhancing idea generation, accelerating R&D, and fostering entrepreneurship by lowering entry barriers.

Workplaces that adopted AI chatbots showed no significant difference in employment, early-career jobs, job churn, or worker mix. Difference-in-differences estimates for earnings and wages are all precisely estimated zeros, ruling out effects larger than 2%.

Calibrated to U.S. data, the model predicts more than threefold productivity improvements alongside a long-run employment loss of 23%, with half (~11.5%) occurring over the initial five-year transition period.

We describe our research with 21 developers at the cutting edge of using AI

Anthropic CEO Dario Amodei warned that up to 50% of entry-level office jobs could be substantially impacted by AI within the next 2-3 years.

Survey of 7,950 decision-makers: 68% of all CS interactions expected to be handled by agentic AI by 2028. 56% expected within 12 months.

25% of global employment falls within occupations potentially exposed to GenAI, with higher shares in high-income countries (34%). Clerical support roles remain the most exposed.

Framework for guiding innovation to increase labor demand; steering technology becomes more desirable the less efficient social safety nets are.

For the empirically plausible case that capital and labor are gross complements, raising wages requires capital-augmenting innovation rather than labor-augmenting.

AI-adopting accountants see 55% increase in weekly client support, 8.5% of time reallocated from data entry to high-value tasks, 12% increase in ledger granularity, and 7.5-day reduction in monthly close time.

AI assistance increases customer service worker productivity by 15% on average. Less experienced workers improve most; AI disseminates best practices from top performers.

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers.

Platform workers in writing, translation, and data labeling report median earnings declines of 15-25% since 2023.

NVIDIA median employee total compensation was $301,233 in FY2025 (ending January 2025), with CEO-to-median pay ratio of 166:1. Total headcount reached 36,000, a 21.6% increase YoY.

Approximately 32% of health employment is classified under potential augmentation, approximately 4.3% of roles are identified as potentially automatable, and the high automation risk category constitutes around 0.6% of the health workforce. Analysis of 55.5 million online job postings.

Over the next decade AI will decimate faculty ranks; TAs will be first casualties, then non-tenured faculty, then tenured positions via attrition. Predicts significant culling of doctoral programs.

With one in three job vacancies having high AI exposure, a significant share of jobs in OECD economies are influenced by the rise of AI.

44% of S&P 500 discussed AI-workforce in Q1 2025, a slight pullback from Q4 2024's record. Discussion maturing: less hype, more specific deployment data.

In a field experiment across 66 firms and 7,137 knowledge workers, the 80% of treated workers who used the AI tool spent two fewer hours on email each week. We do not detect shifts in the quantity or composition of workers' tasks.

31% of AI experts predict AI will lead to fewer teacher jobs in 20 years. Survey of 1,013 AI experts and 5,410 US adults.

Youth (18-24) employment in OECD countries in AI-exposed sectors declined 4.5% YoY. Real starting salaries fell 3-8% across knowledge-work sectors.

75% of knowledge workers use AI at work. 78% bring their own AI tools. Survey of 31,000 workers across 31 countries.

68% of small businesses now use AI regularly (up from 48% in July 2024). 74% say AI makes them more productive. Survey of 2,200+ US businesses.

Augmentation AI fosters new work and raises wages for high-skilled occupations.

In a pre-registered RCT with 776 P&G professionals, individuals with AI matched teams without AI (+0.37 SD improvement). AI broke professional silos.

56% of Americans use AI tools; 28% weekly. Skepticism about AI's effects increased 8 percentage points year over year.

Agentic AI will autonomously resolve 80% of common customer service issues by 2029. Currently only 1.6% of interactions are fully automated end-to-end.

GenAI access leads to increased incorporated entrepreneurship for individuals with higher GenAI exposure -- effect specific to STEM population. Mechanism tests support the augmentation channel.

EY invested $1.4B in AI, reporting that AI-assisted audit procedures now cover 40% of routine testing tasks previously performed by junior staff.

Pearson pivoted to AI-first content delivery, reducing editorial headcount by 18% while growing AI-generated adaptive learning content 300%.

Amazon deployed roughly 800,000 mobile robots across its fulfillment centers by 2024, making it the largest private deployer of mobile robots globally.

In 36 of 49 economies, fewer than 30% of early-stage entrepreneurs consider AI 'very important.' Growing AI Readiness Gap separating entrepreneurs with AI access from those without.

24% decrease in generative AI-exposed skills per firm per quarter in top-quartile automation-exposed jobs. 15% increase in AI-exposed skills for augmentation-prone jobs. Augmentation-prone job postings increased 22% per quarter per firm.

Teleperformance reported full-year 2024 revenue of EUR 10,280 million. Plans to invest up to EUR 100 million in AI-focused partnerships during 2025. Over 200 new AI projects launched in 2024.

AI skills and expertise are highly valued by employers, offering a 23% wage premium, compared to a 13% wage premium for Master's degrees and a 33% premium for PhDs. In science, engineering, and tech jobs, the AI skills premium is 36%. Analysis of over 10 million online job vacancies in the UK between 2018 to 2024.

IBM CEO stated plans to replace ~7,800 back-office roles with AI. Company investing in reskilling remaining workforce for AI-augmented roles.

Fiverr FY2024 marketplace revenue declined 1.3% YoY to $303.1M. Active buyers fell 10% to 3.6M. AI-related categories showed strong growth while traditional low-complexity gig categories contracted.

By 2028, 20% of marketing content will be generated entirely by AI. Creative teams will shrink 15-25% but output will increase.

JPMorgan disclosed deploying LLM-based tools across legal, compliance, and research functions, with estimated productivity gains equivalent to 3,000+ full-time roles.

Shopify reported its AI assistant Sidekick now resolves 40% of merchant support queries without human intervention, up from 12% one year prior.

Posted wages for AI-exposed occupations (admin, data entry, basic accounting) grew 1.2% YoY vs. 4.1% for non-AI-exposed roles, suggesting wage suppression.

BLS projects employment of software developers to increase 17.9 percent between 2023 and 2033. Despite its exposure to GenAI applications, this occupation is unlikely to experience a decline in employment. Customer service representatives projected to decline 5.0 percent. Medical transcriptionists projected to decline 4.7 percent. Paralegals projected to grow 1.2 percent.

Overall employment of computer programmers is projected to decline 9.6 percent from 2023 to 2033.

Software developer job postings 33% below pre-pandemic baseline as of November 2024. AI/ML roles the only growing category.

Alphabet disclosed ongoing workforce restructuring, noting AI tools have reduced need for certain engineering and support roles while increasing demand for AI/ML specialists.

Tenured workers cumulatively lose about 3,800 Euros in wage and salary earnings over five years on average (about 9% of one year's income). Only 0.7% of all workers leave their employers each year due to automation, far below mass layoff rates.

Chegg eliminated 636 employees across two rounds of layoffs (23% in 2024, then additional cuts in early 2025), totaling ~67% of peak workforce. CEO explicitly cited AI as existential threat to homework-help business model.

LinkedIn's AI skills resources hub tracking AI engineering talent growth (130% increase since 2016). 7 out of every 1,000 LinkedIn members globally are AI engineering talent.

Meta's 'Year of Efficiency' resulted in ~21,000 job cuts across 2023. Company cited AI-driven productivity gains enabling smaller teams to maintain output.

Tasks with higher AI exposure subsequently experience reduced labor demand. 1 SD increase in AI exposure reduces relative demand for related skills by 2%.

Workers in AI-exposed occupations face significantly higher unemployment risk.

Workers in AI-exposed occupations face significantly higher unemployment risk, suggesting aggregate job losses may exceed earlier estimates.

Most adversely impacted occupations (business, financial, engineering) experienced a decline of 2% to 2.5% over a five-year period. Reduced demand in exposed occupations is offset by productivity-driven increases in labor demand at AI-adopting firms.

43% of workers in San Jose could see AI shift half or more of tasks, vs only 31% in Las Vegas. High-skill metro areas most exposed — reversing prior automation patterns.

Demand for substitutable skills (writing, translation) decreased 20-50% relative to counterfactual after ChatGPT launch. Short-term (1-3 week) jobs saw sharpest decline. ML programming demand grew 24%; AI chatbot development nearly tripled.

When someone transitions to a job with high exposure to AI, they work 3.5 hours longer per week. Productivity gains captured by firms, not wages.