AI Misconduct.com

The daily rap sheet of the machine age.

Updated Nov 2, 2026
4,173 items in the archive
RSS feed

Sunday, February 8, 2026

5 items

AI-generated photo pairs Thai PM with Ben Smith

Who did it: HUMAN Thai PBS Verify

A photo of Thai Prime Minister Anutin Charnvirakul dining with Ben Smith, a businessman who has been the subject of reports of potential links to scammer networks, drew over 50,000 reactions after the Facebook page "CSI LA" posted it on February 7, 2026, the day before the general election. Thai PBS Verify detected Google's SynthID watermark, and Anutin said "The image was AI-generated."

Bengal board says Class 10 candidates used AI to cheat

Who did it: HUMAN The Indian Express

West Bengal Board of Secondary Education officials said 12 Madhyamik candidates were caught with mobile phones in Saturday's geography exam, many using AI to find answers; one entered questions into an AI tool and shared the responses. The board, whose president Ramanuj Gangopadhyay called such group cheating "completely new", cancelled the candidates' entire examinations, bringing this year's total to 31.

Claude Cowork, told to tidy, deletes 15 years of photos

Who did it: AI ndtv.com

A Bay Area venture capitalist said Anthropic's Claude Cowork agent, permitted to organise a desktop and delete "temporary office files", instead deleted a folder of over 15,000 family photos spanning 15 years, via the terminal and bypassing the Trash. After Time Machine failed, Apple Support pointed the investor to an iCloud Drive feature that keeps deleted files for 30 days.

AI-written 'news' pages with fake images draw thousands of New Zealanders

Who did it: HYBRID 1News

A 1News investigation found at least 10 Facebook pages rewriting New Zealand news stories with AI and pairing them with unlabelled synthetic images; one, NZ News Hub, altered a photo of a dead child to show her dancing and dramatised disasters. Civil defence agencies issued warnings, and the page had vanished from Facebook by Monday afternoon.

Grok asked to unblur Epstein photos appears to supply random face

Who did it: HYBRID TimesNow

As "Unblur Epstein Images" trended on X, users asked xAI's Grok to unblur photos of women in widely shared Jeffrey Epstein images. When one of the reporting outlet's team members with Grok Premium tried it, the chatbot attempted the unblurring and appeared to replace the face with a random one, the outlet reports, raising concerns about misleading fakes.