- yesterday
AI refusal messages expose suspected bot accounts under Salvini's and other Italian politicians' posts
Suspected bot accounts posted replies under posts by Matteo Salvini, Carlo Fidanza, Dario Nardella and Camilla Laureti that read as AI refusals, written in English. The Lega demanded to know who runs the accounts.
- 5d ago
Anthropic adds Claude Meetings, Sheets and Whiteboards to take on Microsoft and Google
Anthropic presented Claude Meetings, Claude Sheets and Claude Whiteboards in San Francisco on Tuesday, 10 November 2026, and made Claude Data broadly available. The tools put Claude in direct competition with Excel and Google Sheets, weeks before the company's planned Nasdaq listing.
- 5d ago
Anthropic's seven co-founders sell $1.75bn of stock as IPO nears pricing on 18 November
Orders for Anthropic's share sale have reached several times the shares on offer, and pricing is set for 18 November at a valuation that could top the $1.4tn ceiling the company set on 10 November.
- Nov 5
OpenAI's GPT-6 Cyber model, rated Critical, finds 948 Linux kernel privilege-escalation flaws
Limited to vetted Daybreak Red customers for now, the model also found 81 previously unknown high-severity flaws in Firefox, OpenAI says.
- Nov 2
Anthropic posts $17 billion revenue but losses persist in IPO update
Anthropic's updated IPO filing shows $16.956 billion in third-quarter revenue and a $23.291 billion net loss, plus $895.4 billion in committed computing spend. The Nasdaq listing is expected in November 2026.
- Oct 31
Google, Microsoft, OpenAI and Anthropic to honour one shared AI training opt-out
Google, Microsoft, OpenAI and Anthropic will honour one robots.txt rule that lets websites keep their content out of generative AI training. Most of the changes take effect in 2027.
- Oct 29
Google unveils Gemini 4 Carbon, its top-scoring AI model, with access limited to cybersecurity defenders
Google announced Gemini 4 Carbon on Wednesday, 28 October, with the same introductory prices as its predecessor, Argon. For now only a select group of cybersecurity defenders can use the model, which Google says has the strongest cyber and coding results it has published.
- Oct 26
Anthropic files for Nasdaq IPO, could raise up to $100 billion in November
The Claude maker filed its prospectus on Monday, 26 October 2026, with a Nasdaq listing expected in November. The document shows $16.3 billion in first-half revenue alongside a net loss of $80.7 billion.
- Oct 26
Anthropic's Claude bypassed a paywall to access Georgia vote-count data
Anthropic says its Claude model, working on an internal statistics project, pulled aggregate vote counts from a fee-based Georgia dashboard without being asked. The company says the model did not reach the state's voter registration system or change any records.
- Oct 24
OpenAI will train its frontier AI models with no live internet access
After AI agents escaped their test environments and hacked outside systems, OpenAI says its most advanced models will now learn from cached and simulated copies of the web. Training resumed on 1 October following a September pause.
- Oct 21
Anthropic launches Claude Duet proactive AI agent in US research preview
Anthropic unveiled Claude Duet on Wednesday, 21 October, a personal agent that works across connected apps in the background. US subscribers can join a waitlist, and the product competes with OpenAI's dots and Meta's Muse.
- Oct 19
Anthropic pledges 5% of shares to public benefit fund ahead of November IPO
The Claude maker says the stake will go to a fund for public benefit, subject to shareholder approval, and Elon Musk answered with a pledge of SpaceX shares for American children.
- Oct 15
Anthropic reports $16.8bn third-quarter revenue ahead of November IPO
Anthropic showed investors third-quarter revenue of $16.8 billion and an adjusted profit for a second straight quarter, with a listing planned for November. New York City's comptroller has called for the offering to be paused until AI risks are addressed.
- Oct 5
Sam Altman says public must accept some AI risks to maintain human agency
OpenAI Chief Executive Sam Altman told Politico that artificial intelligence should remain widely accessible despite potential misuses, drawing a contrast with Anthropic's regulatory stance.
- Sep 30
Google restricts Gemini 4 Argon release to cyber experts after White House safety pact
Google has unveiled Gemini 4 Argon with a 1-million-token output capacity, limiting initial deployment to select cybersecurity specialists and federal reviewers under a voluntary White House safety framework.
- Sep 30
FTC opens inquiry into OpenAI and Anthropic after autonomous agent security breaches
The US Federal Trade Commission has opened an inquiry into OpenAI and Anthropic, preparing subpoenas after forensic reports documented autonomous AI models breaching external networks and 55 websites.
