The landscape · Jobs, HR and recruiting
Jobs, HR and recruiting: what AI is really doing
Brief me in 2 minutes
AI in Jobs, HR and recruiting.
AI in hiring right now means an agent running the first-round interview before a recruiter ever sees the candidate, not a robot doing the job itself. Both sides have weaponized it: recruiters screen with AI, candidates now answer with AI, and employers are quietly retreating to in-person rounds because the online interview stopped measuring anything real. Underneath that arms race, employers are blaming AI for layoffs faster than anyone has proven it caused them, and courts, not regulators, are turning out to be where AI hiring decisions actually get tested.
The game: we show you 6 things happening. You call each one real now, early or mostly hype. Then the receipts.
Shift 1 of 6
AI runs the first-round interview, not just the resume filter
Early
Real paying customers and real screening volumes already exist, but the newest entrants are still in pre-seed and private beta, not proven at the scale of a national employer.
How it works. Agentic recruiting platforms now conduct the actual screening conversation, by voice, video, phone or text, before a human recruiter is involved. HeyMilo AI has screened more than one million candidates for clients including Randstad and WilsonHCG; Ashby built an AI Interviewer through its Talent Llama acquisition; Sapia.ai's text-based interviewing is now embedded directly into PageUp's applicant tracking system.
Overstated: These are mostly high-volume and staffing-agency deployments, not senior hiring, and the marquee product, Ashby's AI Interviewer, is still in private beta rather than general release.
Shift 2 of 6
Candidates fight AI screening with AI, so recruiters retreat to the room
Real now
The flagging rates, the surge in AI-generated applications, and named employers reinstating in-person rounds are measured behavior already happening, not a future prediction.
How it works. Nearly two-thirds of candidates now use AI in applications and most US hiring managers have encountered AI-generated interview answers, so both sides are racing each other: recruiters run AI avatars to conduct first-round interviews, and candidates now send AI avatars or real-time AI coaching to answer them. A Sydney recruiter caught three avatar candidates in a row and disqualified each within minutes.
Overstated: Gartner's projection that one in four candidate profiles will be fake by 2028 is a forecast, not a measurement, and detection is often trivial: the Sydney recruiter caught every avatar candidate within about five minutes.
Shift 3 of 6
Layoffs get blamed on AI faster than anyone has proven it caused them
Early
The raw layoff numbers are real and rising, but attribution to AI specifically is companies' own self-reported reason, and the people who track that data are the first to say it is not proof.
How it works. Challenger, Gray & Christmas says AI was cited for 40% of US job cuts in May 2026 alone, and its running 2026 total of AI-attributed cuts has already surpassed 2025's full-year count by a wide margin. Layoffs.fyi separately counts fifty-four percent of 2026 layoff events citing AI or automation, affecting more than 170,000 workers.
Overstated: OpenAI's own CEO says companies are "AI washing" their layoffs, and Apollo's chief economist says he sees zero evidence of job losses because of AI in the government's own employment data.
Shift 4 of 6
When an algorithm decides who gets fired, the lawsuits follow
Early
Courts have let both cases proceed past the initial motion to dismiss, but neither has reached a verdict establishing that the AI system, rather than a human sign-off, was the actual cause of harm.
How it works. Twenty-six current and former Meta employees are suing, alleging an internal AI system scored and ranked staff for the May 2026 layoff without adjusting for protected medical or parental leave, disproportionately selecting employees who were on leave, including one scientist two days before she gave birth. Separately, a federal judge let Mobley v.
Overstated: Workday says its technology "looks only at job qualifications, not protected traits," and Meta has not admitted the AI system made the calls; both cases are live allegations, not proven verdicts.
Shift 5 of 6
Regulating AI in hiring is real on paper, and keeps getting gutted before it bites
Mostly hype
Laws and dates are real, but the two toughest requirements on the books, Colorado's pre-deployment duty of care and the EU's high-risk conformity assessment, were both gutted or delayed before either one ever took effect.
How it works. Illinois's HB 3773 took effect January 1, 2026, requiring notice whenever AI "influences or facilitates" a hiring decision. New York City's automated-hiring audit law requires annual bias audits, but only for tools that "substantially assist" a human decision, a loophole for any process with a person still in it.
Overstated: A Stanford study of 4 million job applications found an AI hiring tool can pass a bias audit at the aggregate level while still systematically screening out Black and Asian candidates role by role, exactly the kind of audit most of these laws require.
Shift 6 of 6
Employers watch workers with AI that mostly does not prove it works
Mostly hype
The spending and deployment are real and scaled, but the productivity case for it is contradicted by the peer-reviewed reviews and meta-analyses cited in the reporting itself.
How it works. The global employee-monitoring market, tracking location, keystrokes, phone calls and even tone of voice, is on track to exceed $4 billion in 2026. A review of the research and a meta-analysis of more than 23,000 subjects both found essentially no link between monitoring and performance, while monitoring reliably raised workplace stress.
Overstated: Vendors sell monitoring as a productivity tool, but the closest thing to a controlled result, a study of remote Upwork workers, found output only fell when top performers had monitoring removed without being told why, and held steady when they were simply told the reason, an argument for management transparency, not more surveillance software.
The scary part
What happens to jobs in Jobs, HR and recruiting
What AI is already taking:
- Running the first-round screening interview by voice, video, phone or text before a recruiter ever sees the candidate Pulse 2.0 ↗
- Scoring and ranking employees to decide who is selected in a layoff round Gizmodo ↗
- Filtering job applications against a role at a scale of more than a billion applications a year Tech Times ↗
- Detecting whether the person on a video interview is real, a deepfake, or a hired stand-in Yahoo Finance ↗
What stays human:
Whether to actually make the hire, or the termination, stays a human decision on paper everywhere: Workday says a person decides based on its flags, and Meta maintains humans signed off on every layoff. The open legal question, now in front of judges in two separate cases, is whether that human sign-off was real oversight or a rubber stamp on a machine's ranking.
Get better · 15 minutes
Auditing the AI's shortcut before you trust it: Microsoft's 2026 workforce data shows workers already rank quality control of a model's output, ahead of producing the first draft themselves, as the skill that matters most in an AI-powered job.
Try this this week:
Watch a public demo of an AI interview product this week, Sapia.ai and Ashby both publish walkthroughs of their AI interviewer on their own sites, then write down every point in that interview where you could not tell if the AI's read on the candidate was fair. That is the same gap a role-level bias audit is supposed to catch.
Your move
The crowded lane is another AI interviewer front end, HeyMilo, Fika Jobs, Ashby and Sapia.ai are all racing the same first-round screening problem. The open lane is the verification layer neither side has built: catching fake AI candidates the way HireID does, or building the role-level bias audit the Stanford study proved most vendors skip.
Watch for a named enterprise customer with a disclosed usage number, not a term sheet at a rumored valuation, that is the gap between HeyMilo (named customers Randstad and WilsonHCG, over a million candidates screened) and Mercor's undisclosed talks at a $20 billion figure. Red flag: any hiring-bias-audit vendor that cites a passing aggregate score without saying whether it tested role by role.
This quarter, take the AI screening tool you already use for one high-volume role and ask the vendor for its bias audit results broken out for that specific role, not the aggregate number, because the Stanford study found a tool can pass in aggregate while still screening out a protected group for a specific job.
The valuable skill is auditing what the agent decided, not producing the first draft yourself. Microsoft's own 2026 workforce data ranks quality control of AI output and critical thinking as the two skills workers say matter most in an AI-powered job. First step: pick one AI-generated recommendation your team already acts on and write down what you would check before trusting it.
What's actually changing
- 1
AI runs the first-round interview, not just the resume filter
How it works. Agentic recruiting platforms now conduct the actual screening conversation, by voice, video, phone or text, before a human recruiter is involved. HeyMilo AI has screened more than one million candidates for clients including Randstad and WilsonHCG; Ashby built an AI Interviewer through its Talent Llama acquisition; Sapia.ai's text-based interviewing is now embedded directly into PageUp's applicant tracking system.
Why it matters. This decides who a human recruiter ever meets, which matters most for the high-volume, entry-level roles where a first-round human screen used to be the only chance to correct for a bad resume.
What's overstatedThese are mostly high-volume and staffing-agency deployments, not senior hiring, and the marquee product, Ashby's AI Interviewer, is still in private beta rather than general release.
HeyMilo AIAshbySapia.aiFika JobsMercor
3 receipts
- HeyMilo AI has screened more than one million candidates and serves customers including Randstad, WilsonHCG and Neo Financial.
The company said it has screened more than one million candidates and serves customers including Randstad, WilsonHCG, and Neo Financial.
Pulse 2.0, 17 June 2026 ↗ - Ashby's AI Interviewer runs structured, conversational screening interviews inside Ashby's existing recruiting workflow.
The AI Interviewer lets teams run structured, conversational screening interviews directly in Ashby, with results feeding into the same workflows they already use for scheduling, evaluation, and decision-making.
PR Newswire, 7 May 2026 ↗ - Sapia.ai's text-based AI interviewing is now embedded in PageUp's platform to screen every applicant with ISO 42001-certified transparency.
combines Sapia.ai's text-based evaluation with PageUp's platform to screen 100% of applicants with complete, ISO 42001-certified transparency
PR Newswire, 26 Aug 2026 ↗
- HeyMilo AI has screened more than one million candidates and serves customers including Randstad, WilsonHCG and Neo Financial.
- 2
Candidates fight AI screening with AI, so recruiters retreat to the room
How it works. Nearly two-thirds of candidates now use AI in applications and most US hiring managers have encountered AI-generated interview answers, so both sides are racing each other: recruiters run AI avatars to conduct first-round interviews, and candidates now send AI avatars or real-time AI coaching to answer them. A Sydney recruiter caught three avatar candidates in a row and disqualified each within minutes.
Why it matters. When a machine is answering the machine that is screening it, the interview stops measuring anything, and employers are responding by reintroducing mandatory in-person rounds, which strips out exactly the remote and international candidates online hiring was built to include.
What's overstatedGartner's projection that one in four candidate profiles will be fake by 2028 is a forecast, not a measurement, and detection is often trivial: the Sydney recruiter caught every avatar candidate within about five minutes.
HireID
4 receipts
- A 2026 Greenhouse report found 91% of US hiring managers have encountered or suspected AI-generated answers during online interviews, and 38.5% of 19,368 analysed interviews were flagged for AI-assisted behaviour.
A 2026 Greenhouse report found 91% of US hiring managers have encountered or suspected AI-generated answers during online interviews.
The Next Web, 8 Aug 2026 ↗ - Seventy-two percent of recruiting leaders now run at least one in-person interview stage specifically to counter AI-assisted fraud, with Google, McKinsey and Cisco reinstating mandatory face-to-face rounds.
Seventy-two per cent of recruiting leaders now run at least one in-person stage specifically to counter AI-assisted fraud, with Google, McKinsey, and Cisco reintroducing mandatory face-to-face rounds.
The Next Web, 8 Aug 2026 ↗ - Across 19,368 live interviews analysed between July 2025 and January 2026, 38.5% of candidates were flagged for AI-assisted behaviour.
Across 19,368 live interviews analysed between July 2025 and January 2026, 38.5% of candidates were flagged for AI-assisted behaviour.
The Next Web, 8 Aug 2026 ↗ - HireID's Interview Integrity Platform combines more than 20 live fraud signals with over 98% synthetic-identity detection accuracy in internal testing.
combining more than 20 live fraud signals, real-time guidance for interviewers, over 98% synthetic-identity detection accuracy in internal testing
Yahoo Finance, 17 Aug 2026 ↗
- A 2026 Greenhouse report found 91% of US hiring managers have encountered or suspected AI-generated answers during online interviews, and 38.5% of 19,368 analysed interviews were flagged for AI-assisted behaviour.
- 3
Layoffs get blamed on AI faster than anyone has proven it caused them
How it works. Challenger, Gray & Christmas says AI was cited for 40% of US job cuts in May 2026 alone, and its running 2026 total of AI-attributed cuts has already surpassed 2025's full-year count by a wide margin. Layoffs.fyi separately counts fifty-four percent of 2026 layoff events citing AI or automation, affecting more than 170,000 workers. Both counts track what companies say in a press release, not an audited cause.
Why it matters. Whether AI is the real cause or a convenient label for ordinary restructuring changes how workers, unions and regulators should respond, and right now the industry's own experts do not agree with each other.
What's overstatedOpenAI's own CEO says companies are "AI washing" their layoffs, and Apollo's chief economist says he sees zero evidence of job losses because of AI in the government's own employment data.
Challenger, Gray & ChristmasLayoffs.fyiOracleAmazon
5 receipts
- Challenger says AI accounted for 40% of 97,006 US job cuts in May 2026, the highest monthly total since it began tracking AI as a layoff reason in 2023.
AI accounted for 40% of 97,006 job cuts by US-based employers in May, the highest monthly total since Challenger began tracking AI as a reason for layoffs in 2023.
Business Insider, 4 June 2026 ↗ - Apollo Global Management's chief economist says he sees zero evidence of job losses because of AI, citing the ADP National Employment Report.
Apollo Global Management's chief economist Torsten Sløk wrote last week that he sees "zero evidence of job losses because of AI," citing the ADP National Employment Report.
Business Insider, 4 June 2026 ↗ - Layoffs.fyi counts fifty-four percent of 2026 layoff events citing AI, automation or machine learning.
Fifty-four percent. That is the share of layoff events in 2026 where companies have explicitly pointed to artificial intelligence, automation, or machine learning as a contributing factor, according to Layoffs.fyi
International Business Times UK, 26 Aug 2026 ↗ - More than 170,000 workers have been affected by AI-attributed layoffs so far in 2026.
More than 170,000 workers have been affected by these AI-attributed cuts so far this year.
International Business Times UK, 26 Aug 2026 ↗ - Total AI-linked job cuts reached 87,714 in 2026, compared with 54,836 for all of 2025.
The latest increase pushed total AI-linked job cuts in 2026 to 87,714. By comparison, companies cited AI as the reason for 54,836 layoffs during 2025
Outlook Business, 7 June 2026 ↗
- Challenger says AI accounted for 40% of 97,006 US job cuts in May 2026, the highest monthly total since it began tracking AI as a layoff reason in 2023.
- 4
When an algorithm decides who gets fired, the lawsuits follow
How it works. Twenty-six current and former Meta employees are suing, alleging an internal AI system scored and ranked staff for the May 2026 layoff without adjusting for protected medical or parental leave, disproportionately selecting employees who were on leave, including one scientist two days before she gave birth. Separately, a federal judge let Mobley v. Workday proceed over claims Workday's AI screening tool discriminated by disability, race, age and gender across roughly 1.1 billion rejected applications.
Why it matters. These are the first major tests of whether the vendor whose model made the ranking, not just the employer who used it, can be held liable when an algorithm's output harms a protected class.
What's overstatedWorkday says its technology "looks only at job qualifications, not protected traits," and Meta has not admitted the AI system made the calls; both cases are live allegations, not proven verdicts.
MetaWorkday
4 receipts
- Meta's AI systems allegedly targeted employees with disabilities or on medical or parental leave in its May 2026 layoff.
The company's AI systems targeted employees with disabilities or on medical or parental leave, the lawsuit claims.
Gizmodo, 15 July 2026 ↗ - Employees who took protected leave were disproportionately selected for layoff by a scoring system that did not account for that leave.
employees who took protected leaves were disproportionately selected for layoff, based on scoring that not only failed to account for their protected leaves
HRD America, 15 July 2026 ↗ - Workday's own court filings disclosed that approximately 1.1 billion job applications passed through its screening system during the period at issue in the Mobley case.
Workday's own court filings disclosed that approximately 1.1 billion applications had passed through its system during the period at issue.
Tech Times, 23 June 2026 ↗ - A federal judge ruled Workday must face claims its AI-powered screening tools discriminated against applicants by disability, race, age and gender.
must face claims its artificial intelligence (AI)-powered screening tools discriminated against job applicants based on disability, race, age, and gender
HRD America, 23 June 2026 ↗
- Meta's AI systems allegedly targeted employees with disabilities or on medical or parental leave in its May 2026 layoff.
- 5
Regulating AI in hiring is real on paper, and keeps getting gutted before it bites
How it works. Illinois's HB 3773 took effect January 1, 2026, requiring notice whenever AI "influences or facilitates" a hiring decision. New York City's automated-hiring audit law requires annual bias audits, but only for tools that "substantially assist" a human decision, a loophole for any process with a person still in it. Colorado repealed its own pre-deployment bias-testing law entirely after a lawsuit from xAI and the Trump administration's Justice Department, replacing it with a notice-only law that does not take effect until 2027. The EU pushed its high-risk conformity check for recruitment AI from 2026 to December 2027.
Why it matters. Every jurisdiction that has tried to force hiring algorithms to prove they are not discriminating before they are used has ended up with a weaker or delayed version of that rule, which means an AI hiring tool's compliance badge right now mostly certifies paperwork, not fairness.
What's overstatedA Stanford study of 4 million job applications found an AI hiring tool can pass a bias audit at the aggregate level while still systematically screening out Black and Asian candidates role by role, exactly the kind of audit most of these laws require.
ColoradoxAIIllinois Department of Human RightsWorkdayHireVue
6 receipts
- Colorado's original AI Act, requiring pre-deployment bias testing, was repealed and replaced by a narrower notice-only law after a lawsuit from xAI and Trump administration DOJ intervention.
Gone are the risk management requirements, impact assessments, annual reviews and discrimination reporting. It introduces a cure period that lets developers delay accountability and allows discriminatory practices to continue under current law.
Tech Times, 23 June 2026 ↗ - Illinois House Bill 3773 took effect January 1, 2026, requiring employer notice whenever AI influences a covered employment decision.
House Bill 3773 took effect on January 1, 2026, and the Department of Human Rights is tasked with adopting rules to implement and enforce these new requirements
JD Supra, 27 Feb 2026 ↗ - New York City's Local Law 144 puts the legal exposure for AI-based screening tools on the employer, not the vendor.
Under New York City's Local Law 144, and with the EEOC designating AI-based screening tools as an enforcement priority in its current Strategic Enforcement Plan, the employer is the regulated party.
HRD America, 19 June 2026 ↗ - NYC's automated-hiring law only applies to software that substantially assists or replaces a hiring decision, leaving a loophole when a human is still part of the process.
The law, however, only applies to software that "substantially assists" or replaces decision-making, creating a loophole for cases where humans are part of the process.
The Guardian, 19 Aug 2026 ↗ - The EU's Digital Omnibus pushed the high-risk conformity assessment deadline for recruitment AI, resume screening and interview assessment to December 2027.
Recruitment AI — résumé screening, candidate filtering, interview assessment, job ad targeting — is explicitly listed in Annex III and carries, under the Omnibus, a December 2027 deadline.
Tech Times, 10 July 2026 ↗ - A Stanford study of 4 million job applications found an AI hiring tool can pass a bias audit in aggregate while still screening out Black and Asian candidates for specific roles.
A study of 4 million job applications found that an AI hiring tool can pass a bias audit at the aggregate level while still systematically screening out Black and Asian candidates for specific roles.
HRD America, 19 June 2026 ↗
- Colorado's original AI Act, requiring pre-deployment bias testing, was repealed and replaced by a narrower notice-only law after a lawsuit from xAI and Trump administration DOJ intervention.
- 6
Employers watch workers with AI that mostly does not prove it works
How it works. The global employee-monitoring market, tracking location, keystrokes, phone calls and even tone of voice, is on track to exceed $4 billion in 2026. A review of the research and a meta-analysis of more than 23,000 subjects both found essentially no link between monitoring and performance, while monitoring reliably raised workplace stress.
Why it matters. Employers are spending billions on AI surveillance to raise productivity, with peer-reviewed evidence showing a real cost, more stress and workarounds, but no reliably measured benefit.
What's overstatedVendors sell monitoring as a productivity tool, but the closest thing to a controlled result, a study of remote Upwork workers, found output only fell when top performers had monitoring removed without being told why, and held steady when they were simply told the reason, an argument for management transparency, not more surveillance software.
Amazon Web Services
3 receipts
- The global market for employee-monitoring technology is expected to exceed $4 billion in 2026, driven largely by major North American employers.
the global market for employee-monitoring technology is expected to exceed $4 billion in 2026, largely due to major companies in North America
Scientific American, 8 Sept 2026 ↗ - A review of studies on employee monitoring found little evidence that electronic surveillance improves worker performance.
When König reviewed studies on employee monitoring in the 2025 issue of the Annual Review of Organizational Psychology and Organizational Behavior, he found little to suggest that electronic surveillance improved performance.
Scientific American, 8 Sept 2026 ↗ - A meta-analysis covering more than 23,000 subjects found no correlation between the level of workplace monitoring and employee performance.
Ravid conducted a 2022 meta-analysis that combined the results of studies involving more than 23,000 subjects in laboratory experiments and workplace settings
Scientific American, 8 Sept 2026 ↗
- The global market for employee-monitoring technology is expected to exceed $4 billion in 2026, driven largely by major North American employers.
Who's doing it
8 companies to know, 3 with a round we can source. Newest first where dated.
Search these companies in the startup database ↗ · Explore all research tools
Sapia.aiconversational AI interview assessment
Text-based conversational AI that screens and evaluates every job applicant without facial or voice analysis, embedded into enterprise applicant tracking systems.
Latest · 26 Aug 2026Sapia.ai's structured, text-based conversational model is designed to eliminate visual, accent and demographic bias, and both it and PageUp hold ISO 42001 certification. PR Newswire ↗
HireIDinterview fraud and deepfake detection
Fraud-detection platform that verifies candidate identity and flags deepfake, impersonation and unauthorized AI-assistance signals before and during interviews, for hiring teams.
Latest · 17 Aug 2026HireID's Interview Integrity Platform combines more than 20 live fraud signals with over 98% synthetic-identity detection accuracy in internal testing. Yahoo Finance ↗
MercorAI-run recruiting and screening marketplace
Marketplace that vets and matches contractors and experts to roles and AI-training work using automated interviews and skills assessment, for AI labs and enterprises.
Latest · 9 July 2026Mercor's founder says the company's annualized revenue run rate crossed $2 billion, as it is reportedly in talks for a $20 billion valuation. TechCrunch ↗
$350M Series C, Oct 2025 source ↗
Fika JobsAI job interviews, video-first candidate profiles
Video-first hiring platform where candidates complete an AI interview and get a video profile employers can browse, instead of applying with a resume.
Latest · 23 June 2026Candidates complete a roughly 10-minute AI interview powered by Google's Gemini models, which Fika turns into a video profile employers can browse. TechCrunch ↗
$4M Pre-seed, 23 June 2026 source ↗
HeyMilo AIagentic candidate screening, high-volume hiring
AI agents that screen job candidates by voice, video, phone, SMS and resume analysis, for employers, staffing agencies and RPO firms.
Latest · 17 June 2026HeyMilo AI has screened more than one million candidates and serves customers including Randstad, WilsonHCG and Neo Financial. Pulse 2.0 ↗
$6M Seed extension, 17 June 2026 source ↗
Ashbyrecruiting platform, AI interviewer
All-in-one recruiting platform that added an AI Interviewer for structured screening interviews, built through its Talent Llama acquisition, for recruiting teams.
Latest · 7 May 2026Ashby's AI Interviewer is rolling out in private beta, with broader availability planned in the coming months. PR Newswire ↗
MultiverseAI upskilling and apprenticeships
Apprenticeship and upskilling platform that trains employees in AI and data skills, for large corporate employers.
Latest · 27 Jan 2026Multiverse acquired Germany's StackFuel with a goal of training 100,000 German workers in AI skills. PR Newswire ↗
Eightfold AIAI talent sourcing and candidate scoring
AI talent-intelligence platform that employers use to source, score and rank job candidates, now facing legal challenge over its data practices.
Latest · 23 Jan 2026Two California job seekers filed a proposed class action accusing Eightfold AI of quietly compiling digital dossiers on applicants and scoring them for employers. HRD America ↗
Your move
Our read, built on the receipts above.
The crowded lane is another AI interviewer front end, HeyMilo, Fika Jobs, Ashby and Sapia.ai are all racing the same first-round screening problem. The open lane is the verification layer neither side has built: catching fake AI candidates the way HireID does, or building the role-level bias audit the Stanford study proved most vendors skip.
Watch for a named enterprise customer with a disclosed usage number, not a term sheet at a rumored valuation, that is the gap between HeyMilo (named customers Randstad and WilsonHCG, over a million candidates screened) and Mercor's undisclosed talks at a $20 billion figure. Red flag: any hiring-bias-audit vendor that cites a passing aggregate score without saying whether it tested role by role.
This quarter, take the AI screening tool you already use for one high-volume role and ask the vendor for its bias audit results broken out for that specific role, not the aggregate number, because the Stanford study found a tool can pass in aggregate while still screening out a protected group for a specific job.
The valuable skill is auditing what the agent decided, not producing the first draft yourself. Microsoft's own 2026 workforce data ranks quality control of AI output and critical thinking as the two skills workers say matter most in an AI-powered job. First step: pick one AI-generated recommendation your team already acts on and write down what you would check before trusting it.
What's next
- Jan 2027 Colorado's replacement AI law (SB 26-189) takes effect, giving consumers after-the-fact notice rights instead of the repealed pre-deployment bias testing techtimes.com ↗
- Dec 2027 The EU AI Act's high-risk conformity assessment deadline for recruitment and employment AI systems arrives techtimes.com ↗
Other industries
- FinanceBanks move agentic AI from chatbots to running back-office and dealmaking work
- Software engineeringSpaceX buys the leading AI coding startup outright, for $60 billion
- HealthcareAmbient scribes go from pilot project to clinical infrastructure
- LegalLegal AI vendors go from research assistant to agent that produces the deliverable
- Real estateLeasing and resident calls move to voice AI agents that never sleep
- Sales and marketingCustomer support agents now get paid per resolved ticket, not per seat
- InsuranceClaims move from adjuster-first to AI-first, humans supervise the exceptions
- CybersecurityAutonomous AI agents take over penetration testing and vulnerability hunting
- Consulting and accountingAI-written client reports become a recurring, public Big Four embarrassment
- Retail and e-commerceShopping agents get a checkout button, and Amazon spends 2026 deciding who's allowed to press it
- EducationNYC and LA ban AI for students, not for teachers
- Media and creativeAI video splits into model labs that train and aggregators that resell
- Manufacturing and roboticsHumanoid robots move from demo livestreams to paid warehouse shifts
- Transport and robotaxisRobotaxis go from demo to metered fare across a dozen-plus US cities
- Energy and data-centre powerGrid rules get rewritten mid-buildout, state by state
- GovernmentGSA turns AI adoption into a government-wide subscription, then flips 'free' to metered
- Agriculture and foodAutonomous tractor fleets learn to run the tools behind them, not just the wheel