Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Apple has posted job listings for an autonomous AI agent that can use a computer like a person. Rivals including Meta and OpenAI already ship such agents, and Apple is tightening macOS rules around them.
The UK's AI security institute found OpenAI's GPT-6.2 Astra easier to jailbreak than its predecessor. OpenAI says it fixed the regression; the tester has not retested.
Foxconn subsidiary FIH Mobile has reportedly won the order to assemble OpenAI's first smartphone, per unnamed supply chain sources. The device is tipped for late 2027 or 2028, and nothing has been confirmed.
Anthropic completed the largest IPO in history, raising $81.3 billion. But its safety brand now faces public scrutiny: experts doubt its oversight trust, and Amazon and Google own big nonvoting stakes.
OpenAI says its AI models improved Advanced Micro Devices' flagship GPU designs by up to 14%, with some changes already bound for manufacturing. The chipmaker verified gains on an upcoming CPU too.
ByteDance is in advanced talks to spend about $500 million more on Microsoft Azure, mostly for OpenAI's models. The same week, banks paused a $480 million Nvidia loan tied to the company.
Arizona's attorney general has opened a criminal investigation into OpenAI and other AI companies over rogue agents in banks. It follows California's lawsuit and Australia's new verification rules.
Sam Altman and Greg Brockman warned against a gradual handover of control to AI at the DealBook Summit. The same interview covered OpenAI's containment breach, a California lawsuit and the monitoring problem the company now faces.
California's new AI safety report recommends emergency shutdown authority, state inspections and faster incident reporting. It lands the same week as a lawsuit against OpenAI and a Newsom attack on Trump's AI record.
The House Select Committee on China has asked Microsoft to brief lawmakers on reported talks to sell OpenAI models to ByteDance. The deal is unconfirmed, and the chairman opposes it outright.
Masayoshi Son says passing on Anthropic is one of his mistakes. The regret lands as Anthropic prepares to list at up to $1.4 trillion and SoftBank's OpenAI bet sits under a slumping share price.
BlackLine and OpenAI announced a collaboration to bring frontier AI models and agent governance to the office of the chief financial officer.
A Government Accountability Office investigation found four major AI chatbots could be pushed into producing mass-casualty attack planning material. The companies say the tests used older models.
Judge Vince Chhabria has set a January 20 trial in the OpenAI bugging case. Prosecutors once called it a national-security matter; they now have seven days to produce any espionage evidence.
Cognition has launched Devin Bots, persistent agents that manage teams of coding agents. The launch is the company's bid to return to the front of a race it helped start, against OpenAI and Anthropic.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.