AI Misconduct.com

The daily rap sheet of the machine age.

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

Monday, August 17, 2026

9 items

Round Hill Music sues Anthropic and Suno over song lyrics

Who did it: HUMAN Reuters

Round Hill Music sued Anthropic and Suno in California federal court on 17 August, alleging Anthropic used lyrics from at least 500 of its songs, including works by James Brown and The Kinks, to train Claude, and that Suno trained its music generator on them. Round Hill may add "ten thousand or more" compositions, seeking damages that could "potentially approach or conceivably exceed $1 billion."

Sainsbury's pauses Facewatch in one store after wrongful ejection

Who did it: HYBRID the Guardian

Sainsbury's paused Facewatch facial recognition at its East Dulwich store after a customer was wrongly identified as a shoplifter and managers asked him to leave. Sainsbury's apologised and blamed "human error, not the facial recognition technology". Facewatch says it sent "a correct alert" that staff then mishandled. A similar misidentification happened at another south London branch last September.

Massachusetts officer accused of using Flock to track ex

Who did it: HUMAN Boston 25 News

Stow, Massachusetts, police officer Jason Rogers faces possible discipline after a four-week internal affairs investigation found he used state systems multiple times in 2024 and 2025 to get information about a woman he previously dated. She says this included Flock camera searches for her location. The department filed a criminal complaint on July 17 and placed him on unpaid leave.

Tracked book leads 404 Media to Amazon scanning hub

Who did it: HUMAN Adnkronos

404 Media hid an Apple AirTag in a rare book from a 1,000-book Biblio order and tracked it to Amazon's LAS8 warehouse in Las Vegas. According to the outlet, workers there cut off bindings, scan the pages and destroy the books, reportedly to train Amazon's Nova models. Amazon says it buys books "through commercial channels" to improve its products and services.

Fake Kohler–Bullock interview used to push unlicensed crypto platform

Who did it: HUMAN The Sydney Morning Herald

A now-removed fake article showed ABC presenter Alan Kohler interviewing RBA governor Michele Bullock, who storms out over questions about Boulder Sparhaven, a crypto trading platform not licensed in Australia. The interview never happened. ASIC research released alongside says AI is making celebrity-impersonation investment scams more convincing. The regulator took down 19,400 scams last year.

Fake Tatiana Silva sends man a forged ID on WhatsApp

Who did it: HUMAN Public

A man told Ciné Télé Revue that someone posing as former TF1 weather presenter Tatiana Silva contacted him on WhatsApp. The impostor sent a forged identity card and a video greeting him and his daughter, which an AI-detection tool rated 80% authentic. The impostor then asked for money, supposedly for a children's charity, which made him suspicious.

AI picks rugged peak for upset novice hiker; rescue follows

Who did it: AI Hangzhou News

After a row with his wife, a Suzhou man asked an AI app for ideas. It quizzed him on his mood, personality, height and fitness, then recommended Tonggongjian, a largely undeveloped 1,558-metre peak in Lin'an. He fell and was stranded at about 1,300 metres on 12 August. Rescuers say the AI suggested no equipment, and he brought little beyond a folding chair.