MIT Technology Review is reporting that China’s artificial intelligence models are causing significant division among President Donald Trump’s AI advisers. It said that current and former advisers publicly criticized leading U.S. AI companies last weekend, with David Sacks calling Anthropic’s models “lobotomized” and “woke,” and Emil Michael referring to OpenAI’s new head of strategic futures as a “supreme village idiot.” The publication noted this conflict stems from the launch of Kimi, a free, open-source model from Chinese AI company Moonshot, which appears to rival the intelligence of models from OpenAI and Anthropic. The availability of smart, free Chinese models reduces the incentive for companies to pay for U.S. alternatives, creating economic and political challenges for the administration and its AI strategists.
Reuters, Engadget and TechCrunch are reporting that Anthropic’s record $1.5 billion copyright settlement has been approved. Reuters said the plaintiffs alleged Anthropic used pirated works to train its Claude model. Engadget noted it is the largest known copyright payout in history, though TechCrunch added that many authors and creators do not view it as a win.
Axios and Fast Company are reporting that the Trump administration is considering a ban on Chinese AI models, with the launch of Kimi K3 reviving calls for restrictions. Fast Company said officials are divided on these proposals. Separately, the Financial Times and Reuters are reporting that China is considering tighter export controls on its own AI models and chips to prevent Western acquisition of its technology and startups.
Axios and CNBC are reporting that Chris Fall, head of the federal AI Safety Institute (CAISI), resigned after just three months in the role. No reason was given for his departure. The Information and CNBC are reporting that Google is developing a new chip, codenamed “Frozen V2,” to run its Gemini AI models more efficiently, with potential deployment in 2028.
The BBC and South China Morning Post are reporting that the European Union has issued AliExpress a record €550 million fine for unsafe product sales, the largest penalty ever under the Digital Services Act. The South China Morning Post said Alibaba has vowed to appeal the fine.
Full Article: The Download: Chinese AI divides the White House, and a record copyright payout