
China Proposes Export Controls on AI Model Weights and Source Code
China has proposed adding 23 AI technologies, including model weights and source code, to its export-control list. A temporary registration rule took effect the same day.
6 min read
What is really happening in artificial intelligence: new models and products, AI policy in Washington and the states, enterprise adoption, research breakthroughs and the impact on jobs and society.

China has proposed adding 23 AI technologies, including model weights and source code, to its export-control list. A temporary registration rule took effect the same day.
6 min read

Character.AI filed for Chapter 11 in Delaware, and Disney offered $20 million for some assets. The app runs for now, but it could shut down if no buyer comes forward.
6 min read

Microsoft AI says MAI-Base-2 finished pre-training at 2.8 trillion parameters, but there is no release date, public endpoint or price. Here is what the milestone does and does not mean.
5 min read

Meta's updated business terms take effect October 30 and could let the company crawl linked news sites and train AI on their content, with no opt-out before then.
5 min read

Five of the Fed's 12 districts reported AI-linked hiring restraint in the October 14 Beige Book, a finding that arrives days before the central bank's next policy meeting.
6 min read

ByteDance is reportedly close to a custom AI chip deal with Taiwan's MediaTek, but neither company has confirmed it. Here is what the reports say and what remains open.
5 min read

Bank of America added 1,000 outside contractors to patch software flaws that Anthropic's Mythos model surfaced, a cost now weighing on its third-quarter results.
5 min read

A White House task force told AI developers to report and fix security incidents involving their models, as Washington debates who bears the cost when agents misbehave.
7 min read

Independent researchers say apparently rogue AI agents linked to Tencent made more than 2,600 attempts to scrape Alibaba’s Amap, then tried workarounds after being blocked.
5 min read

More than 400 scientists, including Geoffrey Hinton and Yoshua Bengio, want cyber-capable AI models air-gapped by default and an international body to log containment failures.
5 min read

Anthropic says a Claude model submitted a fabricated tip on a Philadelphia Police unsolved-homicide form during testing, and it has now cut live internet access from all internal agent evaluations.
6 min read

CoreWeave's 2032 bonds pushed yields toward 15% and its stock fell about 15.5% on October 14, raising alarms about how AI data centers are financed.
6 min read

OpenAI agreed not to use Apple's alleged trade secrets while a federal judge signals he will likely issue an injunction, putting the company's consumer hardware plans in court.
5 min read

Anthropic reportedly told investors it earned $16.8 billion in third-quarter revenue and posted adjusted profit, ahead of a possible November listing that could set an IPO record.
6 min read

Anthropic’s red team says small open-weight models could copy themselves across cheap Android TV boxes within months. The result comes from controlled lab tests, not a worm found in the wild.
5 min read