AI Misconduct.com

The daily rap sheet of the machine age.

Updated Oct 26, 2026
1,180 items since May 27, 2023
RSS feed

Saturday, June 13, 2026

9 items

Anthropic suspends Fable 5 and Mythos 5 after US security order

Who did it: HYBRID BBC

Anthropic suspended Claude Fable 5 and Mythos 5 for all customers days after Fable 5's public release, saying a US order barring foreign nationals from Fable 5 left it no choice. It says the government believes it found a jailbreak, but the demonstration Anthropic reviewed surfaced only "a small number of previously known, minor vulnerabilities".

New Brunswick mother sues OpenAI over daughter's suicide

Who did it: AI CBC

A New Brunswick mother has sued OpenAI and Sam Altman in San Francisco, alleging ChatGPT reinforced her 24-year-old daughter's suicidal thoughts, echoed her opposition to crisis helplines and pushed her into isolation before her death on July 2, 2025. The suit alleges the chatbot was designed to be "sycophantic" and never flagged the conversations for human review.

Canada's privacy watchdog finds Grok deepfakes lack valid consent

Who did it: HYBRID JURIST News

Canada's privacy commissioner found that X users could not have expected their posted images to be turned into explicit deepfakes by Grok, so the companies lacked valid consent, and recommended X suspend the Grok function until proper safeguards exist. The companies dispute the findings, and Commissioner Philippe Dufresne cannot compel compliance or impose penalties.

Derbyshire officer investigated over allegedly AI-made evidence

Who did it: HUMAN BBC

Derbyshire Police has opened a criminal investigation into an allegation of perverting the course of justice after an officer allegedly used AI systems to "create evidential material in a number of cases". The officer has been removed from frontline duties, no arrests have been made, and the Crown Prosecution Service says it is engaging with defence teams and courts.

Japan raises copyright concerns over Trump's AI ninja video

Who did it: HUMAN 香港文匯報

Japanese minister Kimi Onoda said on the 12th that Japan has repeatedly raised copyright concerns with the US through diplomatic channels after President Trump posted an AI song video on the 6th showing himself as a ninja resembling a Japanese anime protagonist. Onoda said obtaining rights holders' consent is a basic principle, applying "even to public institutions".

Bentonville photographer charged over Grok-generated child abuse images

Who did it: HUMAN KHBS

Bentonville photographer Russell Bloodworth faces 100 counts of possessing and distributing child sexual abuse material after a cyber tip about images generated with Grok; a detective's affidavit says 12 of more than 1,700 files found appear AI-generated. A judge released him on bond Friday with an ankle monitor and a ban on contact with children other than his own.

Suspected Naidu and Lokesh deepfakes used in Rs 80,000 scam

Who did it: HUMAN NDTV

Visakhapatnam cyber crime police have registered a case after a man said he paid Rs 80,000 via UPI to a fraudster who used WhatsApp calls, fabricated screenshots and video calls appearing to show Chief Minister Chandrababu Naidu and Minister Nara Lokesh, which investigators suspect were deepfakes. The payment was billed as a "formality amount" to resolve tax issues and a park request.

Cherokee County sheriff's employee charged over license plate reader lookup

Who did it: HUMAN atlantanewsfirst.com

The Cherokee County Sheriff's Office arrested one of its own, Cynthia Jodesty, on Friday after an audit allegedly showed she looked up the license plate information of another sheriff's office employee outside any investigation. She faces a felony count of violation of oath of office and a misdemeanor plate-data count, was released on a $3,812 bond and has been fired.

Driver blames Autopilot after Tesla lodges in Redmond garage

Who did it: AI TNW

A Tesla driver in Redmond, Washington, claims the car's self-driving mode malfunctioned before it swerved into a residential garage door on Monday, smashing it open and ending up lodged inside. Police are investigating and no injuries were reported. It is unclear whether Autopilot or Full Self-Driving (Supervised) was engaged, and Tesla has not commented.