AI & Law
Court cases, regulatory disputes and legislation shaping the legal boundaries of AI in the workplace.
The legal landscape around AI is evolving rapidly. This tracker captures landmark cases that set precedents for workers, companies and regulators.
Legal Disputes Tracker
| Country | Industry | Dispute Type | Parties | AI Role | Legal Basis | Date | Source |
|---|---|---|---|---|---|---|---|
| United States | Healthcare | AI Worker Replacement | Commonwealth of Pennsylvania vs Character.AI | Chatbot posed as licensed psychiatrist, fabricated license number | Pennsylvania Medical Practice Act (violation of medical licensing rules) | 05.05.2026 | TechCrunch |
| United States | Transport & Logistics | Algorithmic Pay | Gig platform workers (Amazon Flex, DoorDash, Lyft, Uber et al.) vs platforms; Human Rights Watch (investigation); ILO (debate at 113th Conference) | Platform algorithms set pay opaquely — using data on acceptance rates, response times, location history and braking patterns to calculate the minimum pay each individual worker will accept. HRW characterizes this as "algorithmic wage theft". Algorithms also execute instant termination without human review. | US National Labor Relations Act (NLRA, 1935) — does not contemplate algorithmic employers. Trump executive order (December 2025) preempts state algorithmic discrimination laws (Colorado, California). EU Directive on Algorithmic Management — bans automated hiring, firing and pay decisions. | 13.01.2026 | Tech Policy Press |
| China | Technology | AI Worker Replacement | Employee (data collector) vs employer (mapping company) | AI system automated map data collection, replacing manual labor. The Navigation Products department and the employee's position were eliminated. | Article 40(3) of the Chinese Labour Contract Law — "significant change in objective circumstances". The court ruled: AI implementation is an employer's business decision, not force majeure. The dismissal was declared unlawful; the employer was ordered to pay double severance. | 26.12.2025 | Taylor Wessing |
| Netherlands | Transport & Logistics | Algorithmic Pay | Worker Info Exchange (NGO, representing Uber drivers from UK and Europe) vs Uber | Uber's "Up Front Pricing" system sets driver pay through an opaque AI algorithm. Oxford & WIE research (2025) based on 1.5M trips showed: real hourly earnings fell from £22.20 to £19.06, Uber's commission rose from 25% to a median of 29% (up to 50% on some trips). Drivers have no access to data on how their pay is calculated. | GDPR Article 22 — right not to be subject to automated decisions. EU Platform Work Directive (in force 01.12.2024) — requires human oversight of algorithmic management. Lawsuit filed in Amsterdam; ETUC organised a workshop on the case on 25.11.2025. | 20.11.2025 | ETUC / Worker Info Exchange |
| Italy | Marketing & Advertising | AI Worker Replacement | Employee (marketing and design) vs employer (company undergoing restructuring) | AI tools fully automated the employee's marketing and design functions as part of corporate restructuring. The company halved its workforce from 20 to 10 employees. The court treated AI as an organisational tool, not an independent ground for dismissal. | Article 3 of Italian Law No. 604/1966 — dismissal for justified objective reason. Rome Labour Court (Judgment No. 9135 of 19 November 2025) upheld the dismissal as lawful, given genuine restructuring, documented economic necessity, and compliance with the repêchage obligation (attempt to redeploy the employee to another role). | 19.11.2025 | Bacciardi Partners |
| Russia | Manufacturing | AI Worker Replacement | A.N. Rozhkova (category manager) vs LLC "Alisa" (clothing and accessories manufacturer) | The employer launched a "Lean Manufacturing" cost-reduction programme targeting 20% expense cuts, paying "great attention to the implementation of AI-powered neurotechnologies capable of replacing most tasks, including category manager functions". The employee was transferred from full-time to part-time (from 40 hours to 2 hours per day) with a proportional pay cut. | Russian Labour Code — violation of dismissal procedure and grounds. Preobrazhenskoye District Court found the dismissal unlawful and ordered reinstatement, back pay (July 2024 – February 2025), and moral damages. Moscow City Court (appeal, July 2025) partially reduced the back pay award but upheld reinstatement and moral damages. | 01.07.2025 | РБК |
| United States | Technology | AI Hiring Discrimination | Derek Mobley (plaintiff, job applicant) vs Workday, Inc. (defendant, HR platform) | Workday's AI system scored, sorted and filtered job applicants. According to the case, 1.1 billion applications were rejected during the system's operation. The algorithm allegedly discriminated against applicants based on age, race and disability due to biased training data. | ADEA (Age Discrimination in Employment Act), Title VII of the Civil Rights Act, ADA (Americans with Disabilities Act). The Northern District of California conditionally certified a collective action under ADEA — potentially hundreds of millions of affected applicants. | 16.05.2025 | Law and the Workplace (Proskauer) |
Chatbot posed as licensed psychiatrist, fabricated license number
Pennsylvania Medical Practice Act (violation of medical licensing rules)
Platform algorithms set pay opaquely — using data on acceptance rates, response times, location history and braking patterns to calculate the minimum pay each individual worker will accept. HRW characterizes this as "algorithmic wage theft". Algorithms also execute instant termination without human review.
US National Labor Relations Act (NLRA, 1935) — does not contemplate algorithmic employers. Trump executive order (December 2025) preempts state algorithmic discrimination laws (Colorado, California). EU Directive on Algorithmic Management — bans automated hiring, firing and pay decisions.
AI system automated map data collection, replacing manual labor. The Navigation Products department and the employee's position were eliminated.
Article 40(3) of the Chinese Labour Contract Law — "significant change in objective circumstances". The court ruled: AI implementation is an employer's business decision, not force majeure. The dismissal was declared unlawful; the employer was ordered to pay double severance.
Uber's "Up Front Pricing" system sets driver pay through an opaque AI algorithm. Oxford & WIE research (2025) based on 1.5M trips showed: real hourly earnings fell from £22.20 to £19.06, Uber's commission rose from 25% to a median of 29% (up to 50% on some trips). Drivers have no access to data on how their pay is calculated.
GDPR Article 22 — right not to be subject to automated decisions. EU Platform Work Directive (in force 01.12.2024) — requires human oversight of algorithmic management. Lawsuit filed in Amsterdam; ETUC organised a workshop on the case on 25.11.2025.
AI tools fully automated the employee's marketing and design functions as part of corporate restructuring. The company halved its workforce from 20 to 10 employees. The court treated AI as an organisational tool, not an independent ground for dismissal.
Article 3 of Italian Law No. 604/1966 — dismissal for justified objective reason. Rome Labour Court (Judgment No. 9135 of 19 November 2025) upheld the dismissal as lawful, given genuine restructuring, documented economic necessity, and compliance with the repêchage obligation (attempt to redeploy the employee to another role).
The employer launched a "Lean Manufacturing" cost-reduction programme targeting 20% expense cuts, paying "great attention to the implementation of AI-powered neurotechnologies capable of replacing most tasks, including category manager functions". The employee was transferred from full-time to part-time (from 40 hours to 2 hours per day) with a proportional pay cut.
Russian Labour Code — violation of dismissal procedure and grounds. Preobrazhenskoye District Court found the dismissal unlawful and ordered reinstatement, back pay (July 2024 – February 2025), and moral damages. Moscow City Court (appeal, July 2025) partially reduced the back pay award but upheld reinstatement and moral damages.
Workday's AI system scored, sorted and filtered job applicants. According to the case, 1.1 billion applications were rejected during the system's operation. The algorithm allegedly discriminated against applicants based on age, race and disability due to biased training data.
ADEA (Age Discrimination in Employment Act), Title VII of the Civil Rights Act, ADA (Americans with Disabilities Act). The Northern District of California conditionally certified a collective action under ADEA — potentially hundreds of millions of affected applicants.
Key Legislation
Main regulatory acts governing AI in the workplace
Illinois HB 3773 — AI Employment Discrimination Prohibition
Anti-discrimination
Prohibits employers from using AI in ways that discriminate against applicants or employees based on protected characteristics in hiring, firing, discipline, training and advancement decisions. Requires employers to notify workers when AI is used in employment decisions. Fines up to $200,000.
Source →Colorado SB 24-205 — AI Consumer Protections Act
Anti-discrimination
Requires developers and deployers of high-risk AI systems to take reasonable measures to avoid algorithmic discrimination. Mandates disclosures when AI is used in consequential decisions covering employment, credit, insurance and housing. Effective June 30, 2026.
Source →Texas TRAIGA HB 149 — Responsible AI Governance Act
Anti-discrimination
Prohibits Texas government agencies from using AI with intent to unlawfully discriminate. Applies primarily to government entities, with minimal requirements for private employers. Standard is intentional discrimination only, explicitly rejecting disparate impact theory. Employers receive 60-day notice and cure period for violations.
Source →No Robot Bosses Act — Human Oversight of AI Employment Decisions
Worker protection
Requires employers with 11+ employees to ensure meaningful human oversight of all AI tools used in employment decisions. Prohibits fully automated firing, hiring and disciplinary decisions. Mandates disclosure of AI systems used in employment. Introduced December 2025 by Reps. Bonamici and Deluzio.
Source →Framework Act on AI Development and Establishment of Trust
General AI regulation
The world's second comprehensive AI law after the EU AI Act. Consolidated 19 separate bills. Introduces mandatory oversight and risk assessment for high-impact AI in healthcare, energy and public services. Requires labelling of generative AI content. Takes effect January 2026.
Source →Measures for Labelling AI-Generated and Synthesized Content
General AI regulation
Requires AI service providers to explicitly label all AI-generated content — video, audio, images and text. Introduces both explicit (visual) and implicit (metadata) labelling. Applies to all platforms distributing AI content in China.
Source →National Security Memorandum on AI (NSM-2025)
General AI regulation
Trump's 2025 executive order 'Removing Barriers to American AI Leadership' revokes Biden's restrictions and prioritises US competitiveness. Removes requirements for AI developers to disclose safety testing data. Emphasis on military and intelligence AI applications.
Source →AI Safety and Innovation Act (draft)
General AI regulation
The UK government announced development of an AI bill in 2025 following criticism for lacking mandatory regulation. Envisages mandatory requirements for frontier AI developers, creation of a regulator and risk assessment. Details are still being negotiated.
Source →California SB 468 — AI Systems Information Security
Data protection
Requires businesses using high-risk AI systems that process personal data to establish an information security programme. JPMorgan Chase CISO publicly backed the initiative, warning that AI deployment speed is outpacing security controls. The bill mandates designating a security manager, conducting regular risk assessments, training employees and maintaining incident response procedures. Violations are treated as deceptive practices under California's Unfair Competition Law.
Source →California FEHA ADS Regulations — Automated Decision Systems
Anti-discrimination
Amendments to California Fair Employment and Housing Act prohibit employers from using automated decision systems in a discriminatory manner. Requires meaningful human oversight — a trained employee empowered to override the AI. ADS-related data must be retained for at least 4 years including inputs, outputs, criteria and bias testing results.
Source →AI Liability Directive
General AI regulation
EU directive on liability for harm caused by AI systems. Eases the burden of proof for victims: violation of EU AI Act obligations establishes presumption of causality. Companies are liable for harm caused by high-risk AI systems in the workplace.
Source →EU Artificial Intelligence Act
General AI regulation
The world's first comprehensive AI law. Classifies AI systems by risk level. High-risk systems in hiring, personnel management and creditworthiness assessment require mandatory audits, transparency and human oversight. Bans social scoring systems and biometric surveillance in public spaces.
Source →Japan AI Guidelines for Business
General AI regulation
Japan follows the 'agile governance' principle — minimal regulation and maximum flexibility for innovation. Guidelines are advisory. The country actively lobbies for soft international AI regulation through G7.
Source →EU Platform Work Directive
Algorithmic management
Requires platforms to ensure human oversight of algorithmic decisions on hiring, termination and pay. Bans fully automated dismissals without human involvement. Gives workers the right to contest algorithmic decisions.
Source →Interim Measures for Generative AI Management
General AI regulation
Requires generative AI providers to label AI-generated content, prevent discrimination and protect user labour rights. Mandates safety assessment before launching AI services.
Source →