🏛️ Beijing orders Meta to block $2B Manus deal 🏛️
China tightens AI regulation, blocking Meta's biggest cross-border acquisition.

China's National Development and Reform Commission has ordered Meta to reverse its $2 billion acquisition of Chinese AI startup Manus, citing foreign ownership and national security concerns over advanced AI assets. The regulation move landed just weeks before a planned Trump–Xi meeting in Beijing, adding sharp geopolitical tension to the deal. Manus executives are reportedly barred from leaving China during the ongoing probe, and the decision effectively treats frontier AI talent and technology as sovereign assets. This sets a precedent for cross-border AI deal regulation globally. Meta now faces a tight deadline to unwind the transaction, directly impacting its AI strategy at a time when US–China friction over control of foundational AI capabilities continues to intensify.

🎨 Claude adds connectors for Adobe, Blender, Ableton 🎨
Anthropic deepens Claude's integration with professional creative software ecosystems.

Anthropic has launched a major expansion for Claude, adding connectors for Adobe tools, Blender, Autodesk Fusion, Ableton, Splice, SketchUp, Affinity by Canva, and Resolume. These connectors let Claude streamline time-intensive creative tasks, write custom code to extend professional software functionality, and serve as an interactive tutor for mastering complex platforms. The update positions Claude as a deeply embedded assistant across design, 3D modeling, music production, and visual performance workflows — this widens Claude's distribution across the professional creative market. It's a clear differentiation move from competitors still focused primarily on text and code, signaling Anthropic's ambition to capture creative professionals who rely on specialized toolchains daily.

⚖️ Musk's $130B OpenAI trial kicks off ⚖️
The most consequential AI governance and regulation trial begins in court.

Opening statements began in Elon Musk's landmark $130 billion lawsuit against OpenAI, with Musk taking the stand to accuse CEO Sam Altman of "stealing a charity." The suit seeks massive damages, the ouster of Altman and Brockman from the board, and a forced reversal of OpenAI's for-profit conversion. OpenAI's legal team called the suit "sour grapes" from a former backer unhappy with the company's success after his departure. Microsoft's counsel argued Musk only objected once OpenAI became a direct competitor to xAI. The verdict could reshape governance norms for AI organizations. With high-profile figures set to testify and hundreds of pages of private messages entering the public record, this four-week trial is the most consequential in AI history.

INTERESTING TO KNOW

🧪 Talkie: 13B model trained on pre-1931 text 🧪

Researchers have built Talkie, a 13-billion-parameter open-weight language model trained exclusively on data from before 1931 — books, newspapers, public domain records, and cookbooks. Access to a training set free of modern data sidesteps the benchmark contamination problem plaguing frontier models, offering a cleaner testbed for evaluating genuine generalization. Despite never seeing modern code, Talkie can produce working Python by adapting patterns from older logical frameworks, though with lower reliability than contemporary models.

🤝 Google commits up to $40B to Anthropic in performance-based deal 🤝

Google will invest between $10 billion and $40 billion in Anthropic, with the final amount tied to performance milestones. The deal values the AI startup at $350 billion and follows Amazon's recent $5B commitment with similar performance triggers. The funding addresses Anthropic's massive compute gap between supply and demand, positioning it to compete with OpenAI's infrastructure scale while giving Google deeper access to Claude's capabilities.