AI and jobs: what the data shows right now

A dated snapshot of layoffs and hiring research, how many people use AI at work, which AI roles and skills are in demand and what the rules say, with links to every source.

Updated November 1, 2026. We plan to refresh this page monthly. This is research-based, not our own testing or fieldwork: we read the sources linked below. It covers mostly the United States, and many figures come from employers, vendors or models rather than direct measurement, so we say which. Items marked "Reported" come from one or a few outlets, are statements by the people quoted, or are allegations, and we have not confirmed them. This is not career, legal or financial advice.

Where things stand

Employers cite AI for a growing share of job cuts

Challenger, Gray & Christmas counted 43,281 announced U.S. job cuts in September 2026, and employers cited AI for 3,961 of them (about 9%). For January to September, AI was cited for 120,136 of 573,195 cuts (about 21%), the most-cited reason this year, while total announced cuts were down 39% from the same period of 2025. Announced hiring plans were 90,787 in September, the lowest for that month since 2011. The AI figure is the reason employers gave, not a measured cause. [Challenger]

Early-career workers show the clearest gap

Stanford researchers using ADP payroll data found employment for ages 22 to 25 in highly AI-exposed occupations about 19% below what would be expected, with no comparable gap for experienced workers. They describe the data as descriptive, not causal. Indeed Hiring Lab found that as of May 2026, senior postings were up 14.7% over a year and entry-level postings down 7.5%, and presents AI as one possible factor among several. The St. Louis Fed estimated that AI-related job demand explains about 45% of the drop in employment among 18- to 24-year-olds from 2023 to 2025, while saying falling overall openings were the main driver. [Stanford Digital Economy Lab, Indeed Hiring Lab, St. Louis Fed]

Other research finds no clear footprint yet

The Yale Budget Lab's October 14, 2026 update says the mix of occupations is not yet changing in ways that clearly line up with AI's arrival and that AI-usage measures show no link to changes in employment or unemployment. A Chicago Fed working paper covering 2019 to 2024 found employment and wage growth in occupations where AI is highly applicable, which is before much of the 2025-26 activity. Indeed found software development postings up almost 15% since February 2025 while overall postings fell 7%, and cautions that correlation is not causation. [Yale Budget Lab, Chicago Fed working paper 2026-12 on AI applicability and occupations, Indeed Hiring Lab]

Using AI at work is now common, but measures differ

Gallup's May 2026 survey of 22,573 U.S. employed adults found 52% use AI in their role, 30% use it a few times a week or more and 15% use it daily, and 47% said their organization has integrated AI tools. A separate St. Louis Fed tracker found past-week generative AI use for work among employed adults aged 18 to 64 rose from 28.2% to 39.2% between the third quarter of 2024 and the second quarter of 2026, with self-reported time savings of 2.2% of hours. The two surveys ask different questions, so their numbers should not be compared directly. [Gallup, St. Louis Fed FRED Blog]

AI roles and skills in demand

How many postings ask for AI

Lightcast reports that 2.5% of U.S. job postings mentioned AI skills in 2025, up 55% from a year earlier; the figure comes through its analysis of the Stanford AI Index data. Indeed's AI Tracker put postings mentioning AI or generative-AI keywords at 4.2% in December 2025. A keyword mention is not the same as a requirement, and the two sources count differently, so the shares should not be combined. [Lightcast on the Stanford AI Index 2026, Indeed Hiring Lab]

The pay premium, as vendors measure it

PwC's 2026 AI Jobs Barometer, based on more than a billion job ads in 27 countries and territories, reports an average 62% wage premium for workers with AI skills, up from 57%; its page does not explain how the premium is calculated. Lightcast found in 2025 that postings listing AI skills advertised salaries 28% higher, nearly $18,000 a year, than comparable postings without them. Both are consultancy or data-vendor figures, and the Lightcast number is a gap in advertised pay, not wages paid. [PwC, Lightcast]

Roles reported as growing

LinkedIn's Jobs on the Rise 2026 list for the U.S., as relayed by Dice, ranks AI engineer first and AI consultant and strategist second by three-year growth in members' job starts. A SANS tracker, covered by Cybersecurity Dive, counted more than 850 public postings for AI and machine-learning security engineers in mid-September 2026 (median advertised salary $140,000) and about 430 for AI governance, risk and compliance leads (median $146,000). SANS is a security-training vendor, and these are one month of postings. [Dice on LinkedIn Jobs on the Rise 2026, Cybersecurity Dive on a SANS hiring tracker]

Skills and ways in

In Lightcast's 2025 U.S. data, Python was the most common specialized skill in postings (258,674 postings), and postings mentioning agentic AI skills rose from 0.06% to 0.23%. Georgetown's CSET found in 2021, using 2010 to 2020 postings, that employers asked for few AI-specific certifications and that certifications worked mainly as a complement to a degree and experience, which is old evidence. A 2026 survey commissioned by an education provider found 57% of 2,000 U.S. adults see AI certifications as career currency; that measures perception, not hiring. [Lightcast on the Stanford AI Index 2026, Georgetown CSET, Survey for CSU Global]

Reported lately

Newest first. Dates are when the news was reported.

Oct. 30, 2026Reported

Goldman note on workers displaced from AI-exposed jobs

Fortune reported on a Goldman Sachs economist's estimate that workers displaced from AI-exposed jobs since 2025 took about a month longer to find new work and were roughly 10 percentage points more likely to still be unemployed a year later. The "AI-displaced" group is modelled from an exposure score, since no survey asks whether AI caused a job loss. The article also cites BLS data showing 7.4 million workers displaced in 2023 to 2025. [Fortune on a Goldman Sachs note on displaced workers]

Oct. 20, 2026Reported

Greenhouse job-seeker survey

Hiring-software vendor Greenhouse reported that of 1,200 U.S. job seekers it surveyed, 77% used AI in their search in the past year, 48% took an AI interview and 81% were rejected without speaking to anyone at the company; 88% wanted at least one legal requirement for employer transparency about AI in interviews. Fieldwork dates were not stated, and it is a vendor survey. [Greenhouse]

Oct. 14, 2026Reported

Yale Budget Lab update

The Yale Budget Lab's tracker update says the occupational mix is not yet changing in ways that clearly align with AI's introduction and finds no clear AI-related footprint in the labor market. [Yale Budget Lab]

Oct. 1, 2026Reported

Challenger September job cuts report

Challenger counted 43,281 announced cuts in September, down 20% from a year earlier, with AI cited for 3,961, and the lowest September hiring plans since 2011. [Challenger]

Aug. 12, 2026Reported

Stanford updates its early-career analysis

Stanford's Digital Economy Lab updated its "Canaries in the Coal Mine" analysis using ADP payroll data through June 2026. The gap for ages 22 to 25 in highly exposed occupations was about 19%, against 15% in the July 2025 data. The authors say the adjustment looks driven mainly by reduced hiring rather than higher separations, and that the results are descriptive, sensitive to modelling choices and smaller once education is accounted for. [Stanford Digital Economy Lab]

July 2026Reported

EU delays high-risk AI obligations

Law-firm summaries report that the EU's Digital Omnibus on AI, in force since late July 2026, moves the stand-alone high-risk obligations under the AI Act, a category that covers AI used in employment, from August 2, 2026 to December 2, 2027. We read law-firm summaries, not the official text. [CMS]

June 22, 2026Reported

Mobley v. Workday ruling

A federal judge granted in part and denied in part Workday's motion to dismiss in the case alleging its AI screening tools discriminate. California state-law claims and one plaintiff's disability claim survived, while a new race-based claim was dismissed or stricken. These are allegations and procedural rulings, not findings of bias. [Duane Morris]

May 14, 2026Reported

Colorado replaces its AI Act

Governor Polis signed SB26-189, which repeals and replaces Colorado's 2024 AI law with a regime for automated decision-making technology that takes effect January 1, 2027. Employers using it for employment decisions must give notice, explain adverse outcomes and offer human review where reasonable, according to law-firm summaries. The Attorney General enforces, with no new private right of action. [Goodwin]

Dec. 2, 2025Reported

Audit of New York City's AI hiring law

The New York State Comptroller audited enforcement of New York City's Local Law 144, which requires bias audits and notice for automated hiring tools, and called it ineffective for July 2023 to June 2025. The city agency reviewed 32 employer and vendor sites and flagged one likely violation, while auditors found 17 potential violations. The agency agreed to most of the audit's 13 recommendations. [New York State Comptroller]

Open questions

Is AI causing the early-career gap?

Stanford calls its findings descriptive, Indeed lists AI beside remote work, interest rates and post-pandemic corrections, and the St. Louis Fed says fewer openings overall were the main driver. Yale finds no clear footprint. The sources do not agree on how much AI explains. [Stanford Digital Economy Lab, Indeed Hiring Lab, St. Louis Fed, Yale Budget Lab]

How many cuts are really about AI?

Challenger records the reason employers state in announcements. We found no source that measures whether AI was the cause, and Goldman's displacement group is modelled from an exposure score. [Challenger, Fortune on a Goldman Sachs note on displaced workers]

Do AI skills really pay more?

PwC and Lightcast report premiums, but they define them differently, one measures advertised pay, and both are vendors. Salary figures from a single month of postings do not show what people are paid. [PwC, Lightcast, Cybersecurity Dive on a SANS hiring tracker]

Which rules apply to AI in hiring?

The picture is moving: Colorado replaced its law before it took effect, the EU pushed back its high-risk deadline, and New York City's audit found weak enforcement. Rules differ by place, so check the ones where you hire or apply. [Goodwin, CMS, New York State Comptroller]

What training do employers actually offer?

We did not find a reliable 2026 figure from a primary source on how many employers train workers on AI or have an AI-use policy. Gallup's adoption numbers cover use, not training. [Gallup]

If you are job hunting or changing roles

Our own checklist, drawn from the reports above. It is not career or legal advice.

Treat headlines about AI and jobs as one data point: check whether the number is employer-stated, modelled or measured.
If you are early in your career, build and show real work with the tools your target employers use, since postings are shifting toward experienced hires.
Read a posting for what the role does with AI, not only whether the word appears; a keyword mention is not a requirement.
If you use AI to write your application, check every claim yourself and keep it accurate; employers and vendors are also using AI to screen.
Ask employers whether AI is used to screen candidates and how to request human review; some places now require notice.
Treat salary and premium figures as ranges from different definitions, and compare with postings for your own role and city.

Sources

We read the pages above directly or through the outlet named. If a source is wrong or out of date, tell us and we will fix it.