Altman pitches government stake in OpenAI

TechCrunch

Anthropic is in talks with Samsung to develop custom AI chips, joining competitors like OpenAI in seeking alternatives to Nvidia's dominance and addressing chip shortages. The company hasn't determined specific use cases or specifications for the potential chip.

Anthropic

Anthropic restored access to its most powerful AI models after a three-week government-mandated suspension triggered by a cybersecurity jailbreak discovery, announcing new industry-wide jailbreak assessment standards and expanded government pre-deployment testing partnerships to prevent future incidents.

Financial Times

Sam Altman proposes establishing a US-led international AI safety forum and offering the US government a 5% stake in OpenAI, positioning America as the global leader in AI governance while creating an unprecedented public-private partnership model.

Microsoft

Microsoft is investing $2.5 billion to launch Microsoft Frontier Company, deploying 6,000 experts to help customers implement AI at scale while protecting their proprietary data and IP through a model-agnostic platform. Early partnerships with LSEG, Land O'Lakes, Unilever, and Novo Nordisk demonstrate the focus on delivering measurable business outcomes rather than experimentation.

Thinking Machines

Bridgewater's research shows that fine-tuning open-weight models on expert-labeled financial data can surpass frontier LLM performance on judgment-heavy tasks by 29.8% while reducing costs by 13.8x, pointing toward a future of differentiated, domain-specific AI models.