События как стресс-тест рамки

Reuters reported a federal judge sanctioned a managing partner for a false case citation generated with help from an AI legal research tool used by a junior attorney.

Reuters reported Meta's CEO attributed upcoming layoffs to increased capital investment in AI computing infrastructure while saying cuts were not directly due to AI reorganization.

Senate Judiciary Committee advanced the GUARD Act after documented reports of minors being targeted by AI companion chatbots, including sexually explicit and emotionally manipulative interactions. The bill would restrict AI companions for minors and require disclosure that chatbots are non-human.

Reuters reported Alabama's top court dismissed an appeal and sanctioned a lawyer whose filings included fabricated AI-generated citations.

Utah's Medical Licensing Board called for immediate suspension of the state Doctronic AI prescription-renewal pilot, citing patient-safety, oversight, and decision-making concerns around AI-driven refill workflows.

WSJ reported on April 23, 2026 that Mercor, a $10B-valued AI recruiting and model-training contractor platform, was facing at least seven class actions after a breach exposed contractor interview recordings, facial biometric data, screenshots of workers' computers, and other records. Earlier reporting linked the incident to a LiteLLM open-source supply-chain compromise; Meta paused work with the company.

Reliable reporting said Oracle cut thousands of jobs while stepping up AI/datacenter spending; strongest confirmed signal is AI-era resource reallocation, not proven one-for-one substitution.

The FTC announced a settlement under which Air AI and its owners would be banned from marketing business opportunities over allegations they misled consumers.

Anthropic’s March 2026 Economic Index report studied February 2026 Claude usage and learning curves after Opus 4.5/4.6 releases.

Reuters reported draft Russian rules that would allow the government to ban or restrict foreign AI tools if they fail to comply with Russian rules, including data-storage expectations.

Atlassian said it would cut about 1,600 jobs, around 10% of its workforce, while rebalancing toward AI and enterprise sales.

O*NET 30.2 updated 886 occupations and reassigned Job Zones for 325 occupations using a new four-level framework reflecting increased skill demands.

BLS published a Federal Register notice for proposed revisions to the CWS to incorporate updates to digital labor platform work questions in July 2026.

Applicants sued Eightfold AI alleging hidden data collection and AI-generated fit/personality-like scoring without FCRA-style disclosure, consent or dispute mechanisms.

Illinois HB 3773 amendments to the Human Rights Act took effect January 1, 2026, prohibiting discriminatory AI use in employment decisions and requiring notice in certain contexts.

Russia began a voluntary sickness-insurance pilot for self-employed NPD workers from 1 January 2026, improving partial benefit portability for that employment form.

Rep. Valerie Foushee released a congressional report and investigation push requesting clarity from industry leaders about AI’s impact on American jobs.

Russian tax data reported 15.168 million registered self-employed in January–November 2025, up 24.2% from 12.208 million.

The NY State Comptroller found DCWP’s enforcement of Local Law 144 ineffective, including complaint-routing and compliance-review problems.

By December 2025, 35 states + DC had enacted laws or policies regulating student cellphone use in K-12 classrooms (Ballotpedia). 22 of these laws were enacted during 2025 alone (ExcelinEd). Texas, Arizona, Arkansas, Alabama, Tennessee, Kentucky, Georgia, Iowa, Indiana, Idaho, Missouri, Nebraska, Nevada, New Hampshire, New Mexico, North Carolina, Ohio, Oklahoma, Oregon, Vermont, West Virginia, Wisconsin among them.

President Putin called for a national task force to coordinate generative AI development, domestic models and infrastructure, including a target for AI contribution to GDP.

ABC reported that Australian Catholic University had abandoned Turnitin's AI-detection tool in March 2025 after finding it ineffective; students described wrongful or prolonged AI-misconduct accusations and stress around investigations.

Reuters reported NBS data showing China’s 16–24 unemployment rate excluding students reached 18.9% in August 2025, the highest under the revised methodology.

BLS preliminary benchmark data indicated total nonfarm payroll employment for March 2025 would be revised down by 911,000 jobs.

Benioff said Salesforce reduced customer-support headcount from about 9,000 to 5,000 as Agentforce and AI agents handled more work.

Jiang et al. (arXiv:2509.00462) report that in controlled resume-screening experiments LLMs systematically favor resumes generated by the same model over human-written or rival-model resumes, even with content quality controlled. Simulations across 24 occupations show candidates using the same LLM as the evaluator are 23-60% more likely to be shortlisted; effect strongest in business-related fields like sales and accounting. Self-preference can be reduced >50% with self-recognition interventions.

After backlash over an AI-first memo, Duolingo’s CEO clarified that the company had not laid off full-time employees due to AI and that contractor changes were tied to project needs.

Brynjolfsson, Chandar and Chen used high-frequency payroll data and found early-career workers in AI-exposed occupations experienced a relative employment decline after generative-AI adoption.

An unsuccessful applicant filed Harper v. SiriusXM in the Eastern District of Michigan, alleging that SiriusXM's use of an AI-assisted applicant-tracking or screening tool discriminated against him based on race.

California's court system adopted a rule requiring courts to either ban generative AI or establish policies addressing confidentiality, privacy, bias, oversight, transparency and verification.

METR reported an RCT in which experienced developers expected AI to speed work by 24%, but measured task completion took 19% longer with AI on familiar codebases.

Reuters reported Microsoft would cut nearly 4% of its workforce, about 9,100 jobs, while pursuing cost reduction amid heavy AI infrastructure investment.

California's Civil Rights Council announced final approval of regulations clarifying how employment-discrimination law applies to AI, algorithms and automated decision systems.

In Kadrey v. Meta, the court granted Meta's cross-motion for partial summary judgment on fair use on the record presented, while commentary emphasized the ruling's fact-specific limits.

In Bartz v. Anthropic, the court held that using lawfully acquired books for LLM training was fair use, while retaining pirated copies for a library was not protected in the same way.

Texas enacted the Responsible Artificial Intelligence Governance Act (TRAIGA), effective January 1, 2026, creating broad state-level rules for AI systems and prohibited uses, with implications for employers using AI in workplace decisions.

A Guardian investigation reported nearly 7,000 confirmed UK university AI-cheating cases in 2023-24, up from 1.6 to 5.1 cases per 1,000 students year-on-year, while many universities still did not separately track AI misuse.

During the 2025 gaokao, Chinese AI services including Qwen, Doubao, Kimi and Tencent Yuanbao reportedly restricted or paused image-recognition and exam-question-answering functions during exam hours to preserve test fairness.

Ohio State announced an AI Fluency initiative launching in autumn 2025 for first-year students, aiming for every graduate starting with the class of 2029 to be fluent in AI and its application in their field.

Wall Street Journal and Entrepreneur reported (May 2025) blue-book exam booklet sales rising sharply at major US universities: +30% at Texas A&M, +50% at University of Florida, +80% at UC Berkeley over the prior two years. Roaring Spring Paper Products (manufacturer) confirms AI-driven demand. Yale, Princeton, UW-Madison, Merrimack faculty publicly switched from take-home essays to in-class blue-book assessments to neutralize ChatGPT use.

BLS fielded the Contingent Worker Supplement in FY2025/May 2025 after the 2017 and 2023 waves; results were pending as of this corpus update.

Microsoft announced layoffs affecting more than 6,000 employees, around 3% of the workforce, amid a push to streamline management layers and invest heavily in AI infrastructure.

Chegg announced layoffs of about 22% in May 2025; later reporting described further restructuring as AI tools and search changes undermined demand.

Governor Hochul secured statewide bell-to-bell smartphone restrictions in NY K-12 as part of FY2026 budget. No unsanctioned use of phones or internet-enabled personal devices on school grounds for the entire school day. $13.5M for storage solutions. Effective 2025-26 school year.

The U.S. Copyright Office released a pre-publication Part 3 report on generative-AI training, addressing use of copyrighted works to develop generative AI systems.

Klarna’s CEO told Bloomberg that cost-cutting in customer service fueled by AI had gone too far and that the company was recruiting humans to ensure customers can reach people.

IBM CEO Arvind Krishna said AI agents replaced several hundred HR roles, while overall employment increased in programming, sales and other areas.

Honda told US staff they would need to work onsite at least 80% of the week starting 6 October 2025.

Resume Now report (2025): 78% of hiring managers say personalized details signal genuine interest and fit. 62% say AI-generated resumes without personalization often lead to rejection. 79% believe companies should implement formal guidelines on AI-assisted job applications.

A Shopify memo said teams must show why AI cannot do a job before requesting headcount, and that AI capability would factor into reviews.

Stanford AI Index 2025 reported major growth in AI skills in job postings, including generative-AI skill references rising sharply in 2024.

Resume Now survey of 925 US HR workers: 55% report candidates use AI most for resumes/cover letters; 57% noticed uptick in AI-assisted submissions; 90% report increase in low-effort/spam applications driven by AI; 79% believe companies should implement formal guidelines on AI-assisted job applications.

OutRush Wave 4, based on July–November 2024 survey data and nearly 8,600 respondents, reported roughly 8% return between 2023 and 2024 and substantial secondary migration.

Virginia's governor vetoed HB2094, a high-risk AI bill that had passed the legislature and would have required safeguards for AI systems used in consequential decisions including employment.

CAC and other Chinese regulators issued measures requiring explicit and implicit labeling of AI-generated/synthetic content, effective 1 September 2025.

Turnitin announced Clarity, a composition/workflow product giving educators visibility into writing-process evolution and approved AI use. Axios described it as a canvas for students to show their work; Turnitin later positioned it as bringing transparency and responsible AI to the writing process.

Salesforce CEO Marc Benioff said the company would not hire more software engineers in 2025 after AI coding tools reportedly increased productivity.

A Jinan University/Zhaopin report showed flexible-employment job postings rising from 8.4% in 2019 to 15.2% in 2024 and jobseeker share reaching 36.4% in 2024.

Reuters reported Jamie Dimon defended JPMorgan’s five-day RTO policy amid employee complaints, petitions and union-formation inquiries.

The FTC finalized an order requiring DoNotPay to stop making deceptive AI-lawyer claims, provide notice to past subscribers and pay monetary relief.

A Delaware federal judge held on summary judgment that ROSS Intelligence infringed Thomson Reuters' copyrights in Westlaw headnotes used for an AI-powered legal research competitor and rejected fair use as a matter of law.

Anthropic launched the Economic Index to study AI’s effects on labor markets by mapping Claude interactions to occupational/task categories.

Workday announced about 1,750 layoffs, around 8.5% of its workforce, while saying it would prioritize investments in AI and platform development.

Dell CEO Michael Dell told hybrid and remote employees near an office to work onsite five days a week starting 3 March 2025, retiring the hybrid policy.

The U.S. Copyright Office released Part 2 of its Copyright and Artificial Intelligence report on copyrightability, affirming that existing principles require human authorship while allowing protection for sufficiently human creative contributions involving AI assistance.

Stripe cut around 300 employees, mostly in product, engineering and operations, while saying it still planned to reach about 10,000 employees by year-end.

AT&T required many employees to return to offices five days a week from January 2025, concentrating workers in core office hubs.

Business Insider reported Sundar Pichai told staff Google had cut 10% of manager, director and VP roles as part of an efficiency drive; some roles were converted to IC roles.

The FTC approved a final order against Rytr over an AI testimonial and review service that allegedly enabled false and deceptive online reviews.

The Guardian reported on a UK university AI-cheating crisis in which some students admitted using AI, while others described stressful false-accusation or low-evidence academic-integrity hearings. The report also referenced evidence that human and automated detection struggle with AI-authored submissions.

OECD released new Survey of Adult Skills results on 10 December 2024; the Round 1 database covers 31 countries with data collected in 2022–2023.

Directive (EU) 2024/2831 entered into force; member states must transpose it by 2 December 2026.

The FTC proposed action against Evolv Technologies over allegedly deceptive claims about its AI-powered weapons-detection systems, including cancellation rights for certain K-12 school customers.

SafeRent agreed to change its screening practices and pay a multimillion-dollar settlement after allegations that its scoring system discriminated against Black and Hispanic rental applicants and voucher users.

Industry survey of 948 US business leaders found 51% of companies were using AI in hiring (Oct 2024), with 68-70% projected by end-2025. 82% used AI to screen resumes, 64% for assessments, 23% for interviews. 21% allow AI to auto-reject candidates at all stages without human review.

A Senate Permanent Subcommittee on Investigations majority staff report found prior-authorization denials for post-acute care rose at major Medicare Advantage insurers as automated processes were implemented.

Civil-society groups challenged the French CNAF risk-scoring algorithm used to select welfare recipients for investigation, alleging discriminatory targeting of marginalized groups.

The US Department of Labor released AI Best Practices for developers and employers, building on AI and Worker Well-being principles under the AI Executive Order.

The Council adopted rules for platform work that increase algorithmic transparency, require human monitoring of automated systems and give workers the right to contest automated decisions.

The SEC charged Rimar Capital entities and executives over false and misleading statements about purported AI-enabled trading for client accounts and related misrepresentations.

Adobe announced Adobe Content Authenticity, including a Chrome extension available immediately and a free public beta web app planned for Q1 2025 to help creators attach and inspect Content Credentials.

China’s CPC Central Committee and State Council issued guidelines to promote high-quality and full employment, including improved support for flexible employment.

Governor Newsom signed AB 3216 (Phone-Free Schools Act) requiring every California school district, charter school, and county office of education to develop a policy limiting or prohibiting smartphone use during school hours by July 2026.

Texas AG Ken Paxton announced a settlement with Pieces Technologies over allegations of false or misleading claims about the accuracy and safety of healthcare AI products.

Amazon CEO Andy Jassy told staff each s-team organization should increase the ratio of individual contributors to managers by at least 15% by end-Q1 2025 and return to the office five days a week from January 2025.

Cedefop’s 2024 European Skills Index release measures skills-system performance across EU-27+4, using a distance-to-ideal composite approach.

The NO FAKES Act was introduced to protect individuals' voice and visual likeness from unauthorized computer-generated recreations using generative AI and related technologies.

The Verge reported that OpenAI had developed a ChatGPT text-watermarking method and detection tool but had not deployed it, citing concerns including user impact and the fragility of text watermarking compared with alternatives such as metadata.

The EU AI Act entered into force; employment, recruitment, selection, task allocation, monitoring and performance evaluation are among high-risk use cases.

The U.S. Copyright Office released Part 1 of its Copyright and Artificial Intelligence report on digital replicas, addressing voice/visual likeness harms and recommending federal digital-replica legislation.

ABA Formal Opinion 512 addressed lawyers' duties around generative AI, including competence, confidentiality, communication, supervision, candor and fees.

SAG-AFTRA called a strike against video game companies over AI protections, especially voice and performance capture rights; later AP reported a contract ending the strike with AI protections.

The California Supreme Court upheld Proposition 22, preserving app-based driver classification as independent contractors under the statute.

Yandex N.V. finalized the sale of Russian assets in July 2024; the Dutch parent was later renamed Nebius Group, retaining AI-focused international businesses.

Intuit announced roughly 1,800 role eliminations while focusing on AI-powered tax and financial products, framing the move as strategic reallocation.

Turnitin's AI Writing Report guide says scores below 20% are no longer surfaced; instead, an asterisk is shown because percentages between 0 and 19 have a higher incidence of false positives. The guide notes that reports generated before July 8, 2024 may still show numerical scores below 20%.

Scarfe et al. published a PLOS ONE blind real-world test injecting AI-written submissions into a university examination system. The university reported that experienced markers struggled to spot the AI submissions; summaries of the study report that 94% went undetected and AI answers often scored higher than real student answers.

Major record labels including Sony, Universal and Warner sued Suno and Udio, alleging mass copyright infringement through use of sound recordings to train AI music-generation systems.

LAUSD's AI chatbot project with AllHere, marketed as 'Ed', was effectively derailed after AllHere leadership turmoil, furloughs and later fraud charges against the founder; reports describe the district losing millions and shutting down the chatbot soon after launch.

Scarlett Johansson said OpenAI's Sky voice sounded eerily similar to hers after she had declined an invitation to voice the system; OpenAI paused use of Sky and denied intentional imitation.

Colorado enacted SB24-205 requiring reasonable care by developers/deployers of high-risk AI systems including consequential employment decisions.

Google DeepMind announced watermarking methods for AI-generated text and video and described bringing SynthID to key Google products.

TikTok announced it would start automatically labeling AI-generated content uploaded from certain other platforms by using C2PA Content Credentials, describing itself as the first video-sharing platform to implement the technology this way.

EEOC amicus argued Workday could be subject to anti-discrimination law in suit alleging AI screening discriminated by race, age and disability.

On the first anniversary of its detector, Turnitin said the tool had reviewed over 200 million papers since April 2023; over 22 million had at least 20% AI-written content and over 6 million had at least 80% AI-written content, according to Turnitin's model outputs.

OFCCP joined federal agencies in AI guidance and noted that it would analyze federal contractors' AI-based selection procedures under validation and equal-opportunity requirements.

Meta announced that it would start labeling organic AI-generated content in May 2024 and stop removing content solely under its manipulated-video policy in July, moving toward labeling and context.

A federal judge declined to sanction Michael Cohen for mistakenly providing AI-generated fake citations, calling the episode embarrassing but not sanctionable on the facts.

The SEC charged two investment advisers with making false and misleading statements about their purported use of AI; the firms settled and paid $400,000 in civil penalties.

YouTube introduced a Creator Studio tool requiring creators to disclose when realistic content that could be mistaken for a real person, place, scene or event was made with altered or synthetic media, including generative AI.

New Russian rules created interagency commission mechanisms to detect illegal employment and misuse of self-employed contractor status.

Klarna said its AI assistant handled two-thirds of customer-service chats and work equivalent to about 700 full-time agents; later reporting linked AI to shrinking headcount.

The FTC said owners of an AI-branded e-commerce scheme agreed to surrender millions in assets and face bans from selling business opportunities or coaching programs involving ecommerce stores.

British Columbia tribunal found Air Canada liable after its chatbot gave a passenger incorrect information about bereavement fare refunds.

Reuters reported that another New York lawyer faced potential discipline after an AI chatbot invented a case citation in a court filing.

Arizona State University announced a collaboration with OpenAI giving the ASU community access to ChatGPT Enterprise for teaching, learning and discovery, described by ASU as the first higher-education collaboration of its kind.

Duolingo confirmed it cut roughly 10% of its contractor workforce as it used generative AI to streamline content and translation workflows.

The New York Times sued OpenAI and Microsoft alleging unauthorized use of Times journalism to train and operate generative AI systems.

France's CNIL fined Amazon France Logistique €32 million for an excessively intrusive system monitoring warehouse employee activity and performance.

The FTC said Rite Aid failed to implement reasonable procedures and prevent consumer harm when it used facial recognition in hundreds of stores, and the settlement prohibited surveillance use for five years.

A lawsuit alleged Humana used the nH Predict algorithm to deny care in Medicare Advantage, echoing claims around algorithmic estimates overriding clinical judgment.

A proposed class action alleged that UnitedHealth used an AI algorithm to systematically deny elderly patients extended care such as nursing-facility stays.

SAG-AFTRA resolved its 2023 strike with terms including AI and digital-replica protections for performers.

The US executive order on AI directed federal actions on safe and trustworthy AI, including attention to civil rights, worker protections and labor-market impacts.

The 2023 WGA agreement established protections around AI-generated material, disclosure, credit, compensation and training-use disputes.

The Authors Guild and 17 authors filed a class-action lawsuit against OpenAI alleging copyright infringement from use of their fiction works to train GPT models.

Reports described Onclusive's French subsidiary as planning to lay off more than 200 employees while shifting media-monitoring work toward AI-enabled processes.

iTutorGroup agreed to pay $365,000 after EEOC alleged its software rejected female applicants 55+ and male applicants 60+ for remote tutoring roles.

UNESCO published global guidance on generative AI in education and research, aimed at helping countries develop immediate actions, long-term policies and human capacity for a human-centered approach.

Vanderbilt announced that it was disabling Turnitin's AI detection tool, citing concerns about transparency, lack of disable options at launch, and false-positive risk.

Chinese authorities issued interim measures for generative AI services, effective August 15, 2023, with obligations around information security, personal information and content governance.

Seven leading AI companies made voluntary White House commitments that included developing mechanisms such as watermarking to help users identify AI-generated content.

OpenAI updated its AI Text Classifier post to state that, as of July 20, 2023, the classifier was no longer available due to a low rate of accuracy.

Liang et al. published a peer-reviewed Patterns article showing that GPT detectors frequently misclassify non-native English writing as AI-generated.

The Royal Commission into the Robodebt Scheme presented its report after investigating Australia's unlawful automated welfare debt-recovery scheme.

NYC began enforcing AEDT rules requiring recent bias audit publication and candidate/employee notice before use of automated employment decision tools.

A federal judge sanctioned lawyers and their firm after a brief cited fictitious cases generated by ChatGPT in Mata v. Avianca.

After Axel Springer announced cost-cutting at Bild and Welt, reporting and executive statements linked some future editorial role changes to AI-enabled workflows, though the company disputed simple AI-replacement framing.

Turnitin stated that its document-level false-positive rate was below 1% for documents with 20% or more AI writing, but sentence-level false positives were around 4%, and advised educators to use highlighted text as an area of interest to start a conversation rather than draw a conclusion.

BT said it would cut up to 55,000 jobs by 2030; reporting emphasized digitization and AI could eliminate about 10,000 roles.

EEOC released technical assistance on software, algorithms, and AI used in employment selection procedures, emphasizing that employers can be responsible for discriminatory outcomes even when tools are designed or administered by vendors.

A Texas A&M University-Commerce instructor accused a class of using ChatGPT after asking ChatGPT whether it had written student work. Reports said students' grades or graduation status were affected until the claims were revisited.

Chegg warned ChatGPT was pressuring subscriber growth and suspended full-year outlook; shares dropped about 47% in early trading.

IBM's CEO said hiring would pause or slow for back-office roles and about 7,800 jobs could be replaced by AI/automation over several years.

The U.S. Department of Education published a May 2023 report on AI and the future of teaching and learning, recommending human-centered, transparent, equitable approaches to AI in education.

Dropbox announced about 500 layoffs, or 16% of its workforce, with leadership citing slowing growth and the arrival of the AI era while planning to hire new AI-focused talent.

Turnitin activated its AI-writing detection capability for educators amid rapid adoption of ChatGPT and institutional concern about AI-assisted academic work.

Italy's data-protection authority temporarily restricted ChatGPT processing of Italian users' data over privacy and legal-basis concerns.

ProPublica reported that Cigna used PxDx to process large volumes of claim denials, and subsequent legal/regulatory attention challenged the role of automated review in medical-payment decisions.

Sadasivan et al. argued empirically and theoretically that AI-text detectors are unreliable in practical adversarial settings: recursive paraphrasing can reduce detection rates while preserving quality, and watermark-like signatures can be vulnerable to spoofing attacks.

Meta announced it would flatten the organization by removing multiple layers of management and asking many managers to become individual contributors.

The International Baccalaureate stated that it would not ban AI software, arguing that bans are an ineffective response to innovation, and instead required use to align with academic-integrity rules and attribution practices.

OpenAI released an AI Text Classifier and disclosed major limitations: on its English challenge set it identified 26% of AI-written text as likely AI-written and incorrectly labeled 9% of human-written text as AI-written. The same page was later updated to say the classifier was retired on July 20, 2023 due to low accuracy.

NIST released AI RMF 1.0, a voluntary framework for managing risks from designing, developing, deploying or using AI systems.

CNET's AI-generated personal-finance articles drew criticism for errors and corrections; staff later unionized citing editorial independence, layoffs and AI transparency concerns.

Getty Images issued UK proceedings against Stability AI over Stable Diffusion, including claims related to web scraping, training data, outputs, database rights, trademarks and passing off.

China's provisions on deep synthesis services came into effect in January 2023, creating obligations for providers of synthetic text, image, audio and video services.

New York City schools blocked ChatGPT access on school networks in early January 2023 over learning and content concerns, then reversed the block in May 2023 and moved toward teaching-and-learning guidance rather than simple access denial.

Stack Overflow moderators temporarily banned ChatGPT-generated answers, citing high error rates and moderation burden.

A class action alleged Copilot violated open-source licenses and developer rights by training on and generating code from public repositories.

The NLRB General Counsel announced an enforcement approach aimed at protecting workers from intrusive or abusive electronic monitoring and automated management practices.

EEOC and DOJ issued technical assistance warning that AI, algorithms and software used in hiring and employment decisions must comply with the ADA.

The Coalition for Content Provenance and Authenticity released version 1.0 of its technical specification for digital provenance, enabling tamper-evident metadata about content origin and edits.

Spain's rider law amended worker-status rules for delivery-platform workers and is widely described as including algorithmic-transparency obligations through worker representatives.

Italy's data protection authority fined Foodinho, part of the Glovo group, over use of performance algorithms and rider-data practices.

GitHub launched Copilot technical preview, describing it as an AI pair programmer that suggests lines or functions from code context.

HireVue said it was removing facial-expression analysis from its video-interview screening software, while continuing other AI-based assessment features such as language and speech analysis.

Hamburg's data protection authority imposed a €35.3M fine on H&M for data-protection violations involving monitoring of several hundred employees at a service center, including extensive notes about employees' private lives.

Maryland enacted a law prohibiting employers from using certain facial-recognition services during employment interviews unless the applicant consents.

An algorithmic standardization process downgraded many teacher-predicted grades after exams were canceled, triggering backlash and reversal to teacher-assessed grades.

Illinois' Artificial Intelligence Video Interview Act became effective, requiring employers using AI analysis of video interviews to notify applicants, explain the AI evaluation, and obtain consent before evaluation.