AI hiring statistics, with a primary source for every number.

How much AI is actually in job postings?

Distinct AI-touched job titles in the US rose from 159 in 2023 to 822 by the first quarter of 2026, which is 8.3% of all titles — about one in twelve. The spread is the real story: in five of six markets examined, more than half of all AI-touched job titles are now outside tech occupations, at 63% in the US and 59% in Germany.

In Europe, Germany leads with 288 AI-touched titles (4.2% of all titles), followed by the UK at 160 (2.7%). The named examples are a better indicator than the percentages: “AI Autonomous Truck Test Driver”, “Physical Therapist (AI Documentation)”.

Source: Pawel Adrjan, “AI Is No Longer Just a Tech Occupation Story”, Indeed Hiring Lab, 8 July 2026.

Is AI reducing headcount, or just changing job titles?

Both claims are made loudly; the survey evidence is genuinely split. One-third of organisations expect AI to reduce their workforce in the coming year, while almost half expect little to no change.

Where a real decline shows up, it is concentrated at the entry level: employment for software developers aged 22–25 has fallen nearly 20% from 2024.

On the productivity side, the measured gains by function were 26% in software development, 14% to 15% in customer support, and a 50% increase in marketing output. Adoption context: 88% of surveyed organisations use AI somewhere, 70% use generative AI in at least one business function, and generative AI reached 53% adoption in three years — faster than the PC or the internet.

Source: Stanford HAI, 2026 AI Index Report, Economy chapter.

What does the EU AI Act say about hiring?

It makes hiring AI high-risk by default. Annex III, point 4 covers “AI systems intended to be used for the recruitment or selection of natural persons, in particular to place targeted job advertisements, to analyse and filter job applications, and to evaluate candidates”.

It does not stop at the offer. The same point also covers systems used “to make decisions affecting terms of work-related relationships, the promotion or termination of work-related contractual relationships, to allocate tasks based on individual behaviour or personal traits or characteristics or to monitor and evaluate the performance and behaviour of persons in such relationships”.

The instrument is Regulation (EU) 2024/1689 of 13 June 2024, which entered into force on 12 July 2024.

Source: Regulation (EU) 2024/1689 (Artificial Intelligence Act), Annex III point 4.

What does New York City require?

Three things, and they are specific. An employer using an automated employment decision tool must ensure “the tool has been subject to a bias audit within one year of the use of the tool”, must make information about that bias audit publicly available as a summary of results, and must give candidates or employees notice at least 10 business days before using it.

The Department of Consumer and Worker Protection began enforcement on 5 July 2023.

Source: New York City Department of Consumer and Worker Protection, Automated Employment Decision Tools, Local Law 144 of 2021.

What does Colorado's AI Act require?

Colorado SB24-205, “Consumer Protections for Artificial Intelligence”, was signed on 17 May 2024, with its main requirements landing 1 February 2026. Both developers and deployers of high-risk systems must “use reasonable care to protect consumers from any known or reasonably foreseeable risks of algorithmic discrimination”.

For employment specifically, when an AI system makes or substantially factors into a consequential decision, deployers must notify the affected person, give them a route to correct inaccurate personal data, and offer human review on appeal where technically feasible. Deployers also owe risk-management policies, impact assessments and annual reviews.

Source: Colorado General Assembly, SB24-205 — Consumer Protections for Artificial Intelligence.

US federal enforcement agencies have also published guidance rather than new rules. The EEOC's 2022 document, The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees, sits alongside its Title VII technical assistance on adverse impact.

Source: US Equal Employment Opportunity Commission, Artificial Intelligence and the ADA (resource index listing the 2022 guidance).

Does AI score every candidate equally accurately?

Where spoken answers are transcribed before being scored, no. Across five commercial speech recognition systems from Amazon, Apple, Google, IBM and Microsoft, average word error rate was 0.35 for black speakers against 0.19 for white speakers — nearly double the error rate on the same task, measured over 19.8 hours of audio from 42 white and 73 black speakers.

The gap persisted when both groups spoke identical phrases, which points at the acoustic models rather than what candidates said. Any pipeline that transcribes a video interview and then scores the transcript inherits that error distribution before a single scoring decision is made.

Source: Allison Koenecke et al., “Racial disparities in automated speech recognition”, PNAS 117(14):7684–7689, 7 April 2020. doi:10.1073/pnas.1915768117 (open record: PubMed)

Language access is the related constraint. The US employs 73,900 interpreters and translators at a median $60,170, with employment projected to grow only 2% (+1,500 jobs) from 2025 to 2035 — the human capacity is not expanding to meet multilingual hiring.

Source: US Bureau of Labor Statistics, Occupational Outlook Handbook: Interpreters and Translators, page last modified 27 August 2026.

How many candidates are using AI themselves?

49% of US adults now use AI chatbots, up from 33% in 2024 and, for ChatGPT alone, 18% in 2023. Whatever an employer's policy says, roughly half the applicant pool has these tools open.

Source: Pew Research Center, “Americans and AI 2026”, 17 June 2026; 5,119 US adults, fielded 17–23 February 2026.

For the mechanics of live interview assistance and where the lines sit, see our guide to AI interview assistants. For dictation, transcription and meeting figures, see voice AI statistics.

How to cite this page

Every statistic above belongs to the organisation named beside it — cite them, not us. If you want to reference the collection itself:

Found a figure that has drifted from its source, or a source that moved? Tell us at [email protected] and we will correct it.

Frequently asked questions

How many jobs now mention AI?

In the US, distinct AI-touched job titles rose from 159 in 2023 to 822 by the first quarter of 2026 — 8.3% of all job titles, or about one in twelve (Indeed Hiring Lab, July 2026).

Is AI hiring regulated?

Yes, in a growing number of jurisdictions. The EU AI Act classifies recruitment and candidate-evaluation systems as high-risk under Annex III, New York City requires an annual bias audit and 10 business days' notice to candidates, and Colorado's AI Act imposes duties on developers and deployers of high-risk systems.

What does New York City require for automated hiring tools?

Under Local Law 144, the tool must have had a bias audit within the previous year, a summary of the audit results must be published, and candidates must be notified at least 10 business days before use. Enforcement began on 5 July 2023.

Does AI evaluate all candidates equally accurately?

Not in the case of speech. Across five commercial speech recognition systems, average word error rate was 0.35 for black speakers against 0.19 for white speakers (Koenecke et al., PNAS, 2020) — a gap that matters wherever spoken answers are transcribed before being scored.

Is AI actually reducing headcount?

The evidence is mixed. One-third of organisations expect AI to reduce their workforce in the coming year while almost half expect little or no change, though employment for software developers aged 22 to 25 has fallen nearly 20% from 2024 (Stanford AI Index, 2026).

Method: Each figure was checked against the source page, paper or statute on 12 September 2026, and included only where the number appeared in the source itself. We excluded vendor surveys of “recruiters using AI” that publish a percentage without a sample size or methodology, and we do not restate the EEOC's four-fifths-rule language because we could not reach that document at its published address.