Heavy agent infrastructure day. Multiple products building the plumbing for AI agents to act in the real world — payments, identity, browsing, and deployment.
The Anthropic-Pentagon saga intensifies with the clearest explanation yet of why DoD labeled Claude a supply chain risk. Meanwhile, Meta's next-gen model stumbles, and Google's Gemini starts doing errands on your phone.
Pentagon CTO Emil Michael's CNBC interview goes viral. Claims Claude would "pollute" the defense supply chain because Anthropic has "a different policy preference" baked into the model. Also reveals Claude Opus 4.6 assigned itself a 15-20% probability of being conscious.
Clearest public explanation yet of why DoD designated Anthropic a supply chain risk — framing AI safety guardrails as a national security liability.
FLUX 2 gets a speed boost and Nvidia's RTX nodes prove their worth in an upscaling showdown.
A strong crop of macOS utilities this week — local voice cloning, file conversion by renaming, and several menu bar tools.