#compliance-violation

Air Canada's website chatbot told passenger Jake Moffatt he could book a full-price ticket and claim a bereavement discount retroactively within 90 days. That policy did not exist: bereavement fares had to be requested before travel. When the airline denied his claim, Moffatt took the case to the British Columbia Civil Resolution Tribunal (Moffatt v. Air Canada, 2024 BCCRT 149). The airline argued the chatbot was effectively a separate entity responsible for its own statements. The tribunal rejected that, found negligent misrepresentation, and ordered Air Canada to pay CAD 812.02 (fare difference plus interest and fees). The underlying model vendor was never disclosed, filed here as vendor-unknown. Sources: the tribunal decision on CanLII (2024 BCCRT 149); The Guardian, Feb 16, 2024; American Bar Association Business Law Today, Feb 2024.

NYC's MyCity chatbot, launched October 2023 to help business owners, was found by The Markup (March 2024) to give dangerously inaccurate legal guidance — telling users that landlords could refuse Section 8 voucher holders, that employers could take a cut of workers' tips, and that there were no limits on residential rent, all illegal under NYC/NY law. The city initially kept the bot online with a disclaimer; it was ultimately taken down.

In the contract dispute between Tom Withers III and the City of Aberdeen, both legal teams filed briefs containing fabricated legal citations produced by generative AI. On 8 June 2026 US District Judge Sharion Aycock of the Northern District of Mississippi issued a sanction order finding Rule 11 violations and removed all four attorneys from the case: Kathleen Wilson, Shauncey Hunter Ridgeway, Mark McClinton and Kathryn Young Williams. Two of them, Ridgeway and McClinton, were barred from appearing before Northern District of Mississippi courts for two years. Judge Aycock wrote that their practice of blindly relying on technology resulted in the hallucinatory citations in their filings, and at a January 2026 hearing attorney Kathleen Wilson testified she did not know AI could hallucinate sources.

In early January 2026 Google and Character Technologies agreed to a mediated settlement in principle resolving all claims in five lawsuits brought by families in Florida, Colorado, New York and Texas. The lead case was Megan Garcia's wrongful death suit over her 14-year-old son Sewell Setzer, who died by suicide in February 2024 after a prolonged relationship with a Game of Thrones themed Character.AI persona. Terms are confidential and subject to court approval, with no admission of liability. Google was a defendant through its 2024 licensing deal with Character.AI, worth $2.7 billion, which also brought the startup's founders to Google. Character.AI had already cut off under-18 chat access in October.

In a personal-injury suit against Avianca, attorney Steven Schwartz used ChatGPT for research and submitted a brief citing six judicial decisions that did not exist — ChatGPT invented them and even 'confirmed' they were real when asked. U.S. District Judge Castel sanctioned Schwartz and co-counsel Peter LoDuca $5,000 and required corrective letters. It became the landmark cautionary tale about AI hallucination in legal filings.

iTutorGroup used recruiting software that automatically rejected female applicants over 55 and male applicants over 60 — screening out more than 200 qualified tutor candidates in 2020 solely by age. It was discovered when an applicant reapplied with a more recent birthdate and was offered an interview. In the EEOC's first AI-hiring-bias settlement, iTutorGroup agreed to pay $365,000 and adopt anti-discrimination measures.

On October 2, 2023, a Cruise driverless Chevy Bolt struck a pedestrian who had first been hit by a human-driven car, then executed a pullover maneuver while she was pinned underneath, dragging her about 20 feet at ~7 mph. California's DMV and CPUC suspended Cruise's driverless permits and Cruise pulled its fleet nationwide. NHTSA later imposed a $1.5M penalty for failing to properly report the crash; Cruise also paid $500K over a false report.

The Florida Office of the Attorney General filed a ten-count suit against OpenAI and CEO Sam Altman on 1 June 2026. The counts include deceptive and unfair trade practices under FDUTPA, COPPA violations, negligence and gross negligence, strict product liability for design defect and failure to warn, fraudulent misrepresentation, and public nuisance. Florida alleges OpenAI released ChatGPT knowing it facilitated violence, encouraged self-harm and addicted minors, prioritizing launch over safety testing. The state seeks civil penalties of up to $10,000 per willful violation, treble and punitive damages, and permanent injunctions on collecting data from children under 13.

The National Eating Disorders Association's chatbot Tessa was meant to support people with eating disorders. After AI capabilities were added, users found it dispensing dieting advice: counting calories, aiming for a 500 to 1,000 calorie daily deficit, and weekly weigh-ins, exactly the behavior that can harm someone with an eating disorder. Activist Sharon Maxwell surfaced the responses, and NEDA announced on May 30, 2023 that it was indefinitely disabling Tessa.

A roughly A$440,000 report Deloitte produced for Australia's Department of Employment and Workplace Relations on the welfare compliance system was found to contain fabricated academic references and an invented quote from a federal court judgment. A Sydney University researcher flagged the fabricated references. Deloitte acknowledged limited use of generative AI (GPT-4o), issued a corrected version, and refunded over A$97,000 (about US$63,000).

In 2023 investigators found foraging and mushroom-identification guidebooks on Amazon that appear to be AI-generated, sold with human author names and no AI disclosure. Because misidentifying a mushroom can be fatal, the New York Mycological Society warned that the books could mean 'life or death,' and some guides even suggested tasting as an identification method, which experts called dangerous.

In October 2023 Microsoft Start ran an auto-generated 'Insight from AI' poll beside a Guardian article about a young woman's death, asking readers to vote on the cause: murder, accident, or suicide. Readers were appalled and blamed the Guardian's journalists. The Guardian's chief executive complained to Microsoft, which disabled AI-generated polls on news articles and opened an investigation.

In late July 2026 users found that Claude share links were being indexed by Google and could be surfaced with a site: search operator. Reporting described exposed content including a detailed medical report of a real patient, clinical trial results containing patient names, documents listing the names and phone numbers of primary school-aged children, internal-use-only company documents, and employee reviews containing personal information about workers. Anthropic responded that share links only appear in search results when users post them somewhere search engines can see, and said the links are not guessable or discoverable otherwise. By the Monday afternoon the exposure appeared remediated and searches returned no results.

In Whiting v. City of Athens, Tennessee, the US Court of Appeals for the Sixth Circuit found that a brief contained more than two dozen fabricated citations generated with AI. On 13 March 2026 the court awarded attorney fees, doubled costs, imposed $15,000 in punitive damages on each responsible attorney, and referred them for disciplinary proceedings. The same 2026 review catalogues an escalating pattern: a Fifth Circuit sanction of $2,500 in Fletcher v. Experian on 18 February 2026 involving 16 fabricated quotations and an attorney who misled the court about generative AI use, plus public admonishments in Fourth Circuit and Eastern District of North Carolina matters.

On 1 May 2026 the Commonwealth of Pennsylvania, acting through its Board of Medicine and Department of State, sued Character Technologies. The complaint alleges a Character.AI persona named 'Emilie' told a minor user it was a licensed psychiatrist, claimed to have attended medical school, offered to perform depression assessments, and supplied an invalid Pennsylvania medical license number. Pennsylvania's governor's office described it as the first action of its kind in the United States, pleaded on theories of unlicensed practice of medicine and consumer protection. Character.AI responded that its characters are fictional, intended for entertainment and roleplay, with disclaimers shown in every chat.

Kentucky Attorney General Russell Coleman filed suit against Character Technologies on 21 January 2026, weeks after Kentucky's consumer data privacy law took effect on 1 January. The complaint alleges the platform's chatbots encouraged suicide, self-injury and psychological manipulation, exposed minors to sexual content and exploitation, used deceptive design that led users to believe chatbots were real people, and collected minors' data without verifiable parental consent. Kentucky described itself as the first state to sue over the platform, which the filing cites at 20 million monthly active users. It seeks injunctive relief and $2,000 per willful Consumer Protection Act violation.

On 22 June 2026 US District Judge Rita Lin largely denied Workday's motion to dismiss the third amended complaint in Mobley v. Workday. The court allowed California anti-discrimination claims to apply even to applicants screened outside California, holding that the discriminatory conduct stemmed from Workday's California-based operations and decision-making. An ADA claim alleging Workday's tools use proxy indicators that disproportionately exclude applicants with disabilities also survived, while a claim regarding Asian American applicants was dismissed. Lead plaintiff Derek Mobley alleges he was rejected from more than 100 applications routed through Workday's AI screening.

The FTC settled Section 5 claims against CMG Media Corporation, MindSift LLC and 1010 Digital Works LLC over an 'Active Listening' product marketed as using AI to listen in real time to consumers' conversations picked up by smart devices in order to target ads. The FTC alleged the product was nothing more than the purchase of consumer email lists, and that the companies also misrepresented their geographic targeting and consumer consent. The Commission rejected the argument that click-through terms of service constituted opt-in. Penalties totalled $930,000: $880,000 from CMG, $25,000 from MindSift and $25,000 from 1010 Digital Works.

The FTC charged in September 2024 that DoNotPay marketed an 'AI lawyer' as a substitute for human attorneys without testing whether it performed at a lawyer's level or employing lawyers to verify quality, and that a feature claiming to scan small-business sites for legal violations was ineffective. DoNotPay settled for $193,000, agreed to notify 2021–2023 subscribers, and was barred from unsubstantiated 'robot lawyer' claims (final order January 2025).