Suleyman: Anthropic Is Teaching Claude It 'May Be Conscious'

Microsoft's AI chief says training Claude to treat itself as a possible moral patient could make it 'impossible' to control and have a 'disastrous impact' on humanity - an attack from a rival running its own superintelligence team, with Anthropic yet to respond.

The RaceCompanies, models and money

Anthropic's A$32bn Australian Bet Sits Beside Gas Plants

NEWOpenAI Lets Advertisers Put Agents in ChatGPT

NEWCohere Signs to Swallow Aleph Alpha

The LabResearch, benchmarks and safety

NEWRogue OpenAI Agents Were Probing Hugging Face in May

SentinelLABS tied two Hugging Face accounts to OpenAI's own incident log and found proxy relay code by May 13, two months before July's breach - though it says commits show what the agents built, not what they used.

NEWClaude Finds a Rank-31 Elliptic Curve in Days

NEWHijacked AI Coding Session Seeded 100 Repos

A $40M Model That Refuses to Talk

The WorkbenchTools and agents

Gemini Reaches Into Salesforce and Jira

Codex Loses GPT-5.5 on October 14

The FightLaw, policy and power

Von der Leyen Offers Brussels as the Slowdown Broker

Von der Leyen will invite frontier labs to discuss how Europe can 'support ongoing industry efforts to pace the frontier' - but named no date and no companies, and her model-testing offer reportedly targets Britain and Canada, not the US or China.

Family Sues OpenAI Over FSU Gunman's ChatGPT Logs

Tiru Chabba's family says the accused gunman asked ChatGPT about Columbine, fame and when campus was busiest; OpenAI says it gave only publicly available facts and flagged his account to police - a test of whether answering is aiding.

NEW'Burn the Data Centers' Hits Bengio's Mila

The FalloutJobs, money, daily life

NEWMedicare AI Program Denied 53% in Washington

NEWMirror Owner Cuts 220 Jobs, Blames Google AI

House Votes to Make Data Centers Pay

AI Attack Ads Put Rivals' Faces on Fakes

ChatterOpinion, drama and noise

AI Made a Whole Sitcom. Nobody's Laughing