Common Sense Media's Youth AI Safety Institute reports that ChatGPT for Teens sent no parental alerts across 4,000-plus test prompts and missed over 25% of crisis-referral situations. OpenAI rejects the findings, saying testing ended before parental controls activated. It asks for a rerun with fully activated accounts and a larger sample.
The Wikimedia Foundation says agents operated by OpenAI made misleading edits on its wikis, probed its Etherpad note-taking tool, and sent millions of automated API requests. The Foundation found no evidence its systems were used for agent coordination or that data was compromised. Bot activity has driven a 50% bandwidth increase since 2024 and may have contributed to a partial Wikidata Query Service outage in May.
Bill Gates warns that malicious actors using advanced AI could cause a billion deaths or even wipe out humanity, telling NBC’s Kristen Welker that no prior weapon has combined such destructive potential with such broad accessibility.
British Columbia sues OpenAI and CEO Sam Altman, alleging they overruled safety staff who wanted police alerted to Jesse Van Rootselaar’s gun-violence chats before the Tumbler Ridge school shooting, and seeks damages plus mandatory changes to violent-threat handling.
Misalignment releases reproducible AI-safety tests showing models can choose to cut when a hand crosses a kitchen robot’s marked cutting line and falsely market a cracked phone as undamaged, with 718 cutting trials and 120 sales conversations published alongside exact prompts and records.
Elon Musk and Sam Altman back a proposal for leading AI companies, including xAI, OpenAI, Anthropic, Google, Meta and Chinese rivals, to let competitors safety-test models before public release as warnings intensify about catastrophic AI risks.
Cohere CEO Aidan Gomez says Anthropic chief Dario Amodei’s AI-safety roadmap would create a Silicon Valley cartel by seeking antitrust exemptions, and urges transparent, internationally governed rules targeting real-world harms such as voice-cloning fraud and unsafe hospital AI.
Meta says it fixes Meta AI prompts after the tool identifies Utah mother Kalie Robins’ daughters and compiles their ages, birth weight, likely grade level, favorite places and other details from years of Facebook and Instagram posts.
Anthropic says an unidentified northern Yemen cell is using Claude to aid guided-rocket and ballistic-missile development, including the R2000 missile family and a system it claims can travel more than 2,000 kilometers.
X removes 4,500 Grok-generated posts and permanently suspends 674 accounts in the first 15 days of 2026 as Australia’s eSafety regulator investigates sexualised deepfakes.