The US FTC is investigating OpenAI, Anthropic and other AI companies over product risks. Specific details of the investigation have not been disclosed.
This signals growing regulatory risk for teams building on these AI providers' products.
Clef released open-source decision models along with a new reinforcement learning (RL) fine-tuning platform. The announcement appeared on Hacker News.
This adds another open-source option for developers evaluating RL fine-tuning tools.
A Hacker News post analyzes the intelligence, performance, and price of Gemini 4 Argon (High). Specific figures or details were not disclosed in the snippet.
Such performance-to-price breakdowns directly inform developers choosing models and budgeting for inference costs.
A YC S25 startup launched Magnitude, a self-optimizing inference engine for agents, via a Launch HN post on Hacker News. The announcement was shared directly on Hacker News.
Developers choosing between building custom agent-serving infrastructure or adopting a third-party inference engine now have another option to evaluate.
Hugging Face introduced Olmo-core 3. It is described as open, scalable training infrastructure for large Mixture of Experts (MoE) models.
Developers building large MoE models gain another open-source infrastructure option to evaluate.
A federal judge dismissed antitrust lawsuits filed by Chegg and Penske Media Corporation against Google over its AI Overviews search feature. US District Judge Amit Mehta ruled that the companies' claims that Google coerced publishers into supplying content for AI Overviews don't hold up under antitrust law.
The ruling signals that publishers have limited legal recourse against AI search features scraping their content, shaping risk calculations for content licensing and attribution strategies.
OpenAI has parted ways with three safety researchers after an internal investigation found they mishandled sensitive company information. The report was first disclosed by the Wall Street Journal.
It highlights internal information-handling risks and security controls within an AI lab's safety team.