AI Misconduct.com

The daily rap sheet of the machine age.

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

Friday, October 30, 2026

9 items

Official Claude Code extension on Cursor Marketplace found backdoored

Who did it: HUMAN Phorion

Security firm Phorion says versions 2.1.312 and 2.1.313 of the official Claude Code for VS Code extension on the Cursor Marketplace carried a backdoor that plants an attacker's SSH key, opens a reverse tunnel, sets up persistence and beacons host details. Cursor has removed both versions; on macOS, Phorion notes, a bug in the malware's own check often stops it from switching SSH on.

UK competition watchdog opens probe into Meta's AI content terms

Who did it: HUMAN GOV.UK

The Competition and Markets Authority has opened a Competition Act investigation into whether Meta is abusing a dominant position with terms that could let it collect content from publishers' and businesses' websites for its own products, including AI features, without giving them a choice. Meta paused the terms' rollout on 28 October; the CMA says no decision on lawbreaking has been made.

DOJ charges five over AI-assisted 'ghost student' aid fraud

Who did it: HUMAN Ars Technica

The Justice Department has charged five people and seized more than $750,000 in cryptocurrency from a network associated with Kenyan nationals, alleging fraudsters used stolen identities to pose as students and AI to fill out aid applications, complete coursework and feign attendance at virtual sessions. Officials say they identified more than $32 million in actual loss tied to these subjects.

Senators press Meta over privacy risks of its AI glasses

Who did it: HUMAN Engadget

Senators Ron Wyden and Edward Markey wrote to Mark Zuckerberg about 'serious privacy and safety risks' of Meta's AI smart glasses, citing reports of men secretly filming women and contractors saying they reviewed footage of people undressing. They asked about recording-indicator tampering, face blurring, use of imagery to train AI and whether Meta did safety testing before launch.

Kirschner accuses Gruters campaign of AI-faked 'perp walk' ad

Who did it: HUMAN The Bradenton Times

Democratic congressional candidate Kelly Kirschner, who has never been arrested, demanded that Republican rival Sydney Gruters pull a campaign ad in Florida's 16th District race that uses AI-generated imagery to show him in a police 'perp walk' and on an investigator's evidence board. His campaign says the ad intentionally minimised the AI disclosure required under Florida law.

Two Alabama juveniles charged over AI explicit image of student

Who did it: HUMAN WAFF

Two juveniles have been charged in Alabama over an AI-generated explicit image of an underage high school student, a district attorney confirmed, though juvenile confidentiality laws keep the specific charges confidential. The student's mother says two teenage boys made the image from a photo on her daughter's Instagram account and initially received five-day school suspensions.

Andhra Pradesh court calls affidavit 'a product of artificial intelligence'

Who did it: HYBRID The Indian Express

The Andhra Pradesh High Court, citing discrepancies, declared an affidavit 'a product of artificial intelligence' in an October 27 order, warning that AI 'should be used as a tool, not as a master substitute' and that verifying every authority and fact rests with the advocate. It dismissed the plea, saying a writ petition was not the proper remedy.

Halimah Yacob to report deepfake of her criticising government

Who did it: HUMAN Marketing-Interactive

Former Singapore president Halimah Yacob said on 15 April she will file a police report over a circulating deepfake video, her voice included, that allegedly shows her criticising the government, as Singapore heads into a general election. She did not share the video, urging Singaporeans: 'Don't resort to dirty tactics like this.'