AI Misconduct.com

The daily rap sheet of the machine age.

Updated Oct 24, 2026
1,041 items since May 27, 2023
RSS feed

Thursday, July 2, 2026

10 items

Indian top court voids tribunal rulings resting on fake AI precedents

Who did it: HYBRID The Indian Express

A Supreme Court bench of Justices P S Narasimha and Alok Aradhe set aside an August 28, 2024 NCLT order and a September 11, 2025 NCLAT order that cited non-existent judgments and AI-generated paragraphs. It likened hallucinated precedents to a methyl isocyanate release and asked the Bar Council of India to set norms and disciplinary action.

MeetingTV sues Palo Alto over allegedly hallucinated espionage report

Who did it: HYBRID theregister

MeetingTV has sued Palo Alto Networks, its newly acquired Koi Security and Koi researchers, alleging Koi's AI-driven 'Wings' platform hallucinated findings in a December 30 blog that labelled its Zoomcorder product a front for a Chinese espionage operation, which MeetingTV says led security firms worldwide to block its domains. Palo Alto expects the dispute to be resolved through legal process.

Dutch probe ties AI-written EU tobacco submissions to Philip Morris

Who did it: HYBRID Génération sans tabac

A KRO-NCRV and NOS analysis of more than 65,000 responses to an EU tobacco-directive consultation found 22,500 entirely AI-generated, mostly from a Philip Morris International tool promoted as 'Your Voice Your Choice' that named the company only in small print, offered mostly anti-regulation arguments and, in at least 50 cases, added arguments users hadn't selected.

Taipei police arrest 12 more in AI-voiced romance scam

Who did it: HYBRID 三立新聞網 setn.com

Taipei prosecutors and police arrested 12 more people, including eight hostesses and an AI engineer, in a second sweep against a ring allegedly led by Huang Chien-hao that, police say, since 2019 used AI-simulated female voices to court men online, then steered them into buying goods through fake websites, netting several million New Taiwan dollars from more than 10 victims.

Anthropic says Alibaba-linked accounts ran 28.8 million Claude exchanges

Who did it: HUMAN Memeburn

In a June 10 letter to US senators, Anthropic alleged that operators tied to Alibaba's Qwen lab used nearly 25,000 fake accounts to run 28.8 million exchanges with Claude between April 22 and June 5, 2026, targeting coding, reasoning and cybersecurity skills to copy its capabilities. Alibaba denies wrongdoing; Anthropic wants federal rules on API misuse.

California appeals court says pro se appellant failed to show cause against sanctions over fake citations

Who did it: HYBRID CA California (4d)

California's Fourth Appellate District dismissed Trina Patterson's appeal against Nuvision Credit Union after the parties settled, but found Patterson failed to show cause against sanctions over an opening brief citing nonexistent cases, misrepresenting others and quoting passages absent from cited authorities, hallmarks of AI hallucination, the court said.

Renters sue UDR over RealPage pricing under San Diego algorithm ban

Who did it: HUMAN classaction.org

A proposed class action filed July 2, 2026, alleges UDR used RealPage algorithmic tools, including YieldStar, to set San Diego rents and occupancy levels from nonpublic competitor data, violating a city ordinance in force since June 21, 2025. The complaint cites UDR's admission, in District of Columbia litigation, that it used YieldStar as "one tool among others".

Judge blocks Sony from adding 30,442 recordings to Udio suit

Who did it: HUMAN Music Business Worldwide

US District Judge Alvin K. Hellerstein denied Sony Music's request to add 30,442 sound recordings, which Sony said it found in Udio's training data during discovery, to its copyright suit against the AI music platform, saying it would materially alter the case near the close of discovery. The suit stays at 333 works; Udio argues its training is fair use.

Trump posts AI clip curing celebrity critics of 'TDS'

Who did it: HUMAN Clarín

President Donald Trump posted a roughly 90-second AI-generated video casting himself as a doctor treating 'Trump Derangement Syndrome', featuring synthetic versions of Robert De Niro, Julia Roberts, Whoopi Goldberg, Rosie O'Donnell, Edward Norton and John Leguizamo as patients, before prescribing prayer and Diet Coke. None has said they consented; deputy chief of staff Dan Scavino reposted it on X.

Wells Fargo plaintiff told to explain withdrawn nonexistent citations

Who did it: HYBRID C.D. California

After Wells Fargo said a pro se plaintiff's remand motion contained 'AI-hallucinated citations', the plaintiff withdrew several authorities that do not exist or do not support his arguments, without saying whether AI was used. Judge John F. Walter ordered him to show cause by July 9, 2026 why he should not face $1,000 sanctions and/or dismissal.