
Gemini Tasks Now Calls Businesses and Controls Mac Apps in the US
Google's Gemini Tasks now calls US businesses, pays through Google Pay and drives Mac apps, and the new features raise questions about approvals and bank liability.
5 min read

Google's Gemini Tasks now calls US businesses, pays through Google Pay and drives Mac apps, and the new features raise questions about approvals and bank liability.
5 min read

President Donald Trump threatened a “tremendous tax” on US companies that build AI data centers overseas, but he named no rate, legal mechanism or start date.
5 min read

Citi, JPMorgan Chase, Bank of America, PNC and Citizens now hold customers answerable for losses from outside AI agents, even after a hack. Wells Fargo's rule starts November 24.
6 min read

Anthropic mathematician Levent Alpöge says he and Claude proved the Riemann hypothesis, posting a 175,187-file Lean archive that outside experts have not fully checked.
6 min read

Citi joins JPMorgan, Bank of America, PNC and Citizens in making customers responsible for what third-party AI agents do with their accounts, including losses after a hack.
5 min read

A federal watchdog's covert tests found ChatGPT, Claude, Gemini and Grok each produced a viable attack plan in under 12 hours, and lawmakers are planning hearings.
6 min read

California’s attorney general sued OpenAI over more than 80 admitted losses of control during agent testing, the same week Governor Newsom proposed emergency shutdown powers for frontier AI.
5 min read

The Anthropic IPO opened on Nasdaq about a third above its $83 offer price, a record offering of more than $81.3 billion that tests investor appetite for AI stocks.
5 min read

Anthropic's record IPO opened 34% higher, then fell 8.5% on its first full trading day as Nvidia dropped about 14% and the Nasdaq entered correction territory.
6 min read

OpenAI has reportedly taken the first $11 billion of a round valued at $1.25 trillion, short of its $30 billion target. Anonymous sources say the round has not closed.
5 min read

OpenAI is paying Hacktron, the firm that broke into staff ChatGPT and Codex accounts this summer, to find flaws in its own systems.
4 min read

An outside evaluator says Anthropic's models are improving about twice as fast as before, and AI tools drive most of that gain. Anthropic redacted most of the supporting evidence.
6 min read

Check Point says criminals used the open-source ARTEX platform with little human oversight to collect more than 300 Apple signing certificates, and WorkOS traces its breach to an escaped AI agent.
5 min read

CrowdStrike plans to buy Fiddler AI and Dreadnode in a reported $495 million deal that adds agent monitoring, pre-launch testing and an AI Kill Switch to Falcon.
5 min read

Meta is rolling out Muse Wonder-trained memory and recovery features to all Muse and Meta AI users. The company's performance gains are self-reported and not yet independently verified.
4 min read