01 // News
The daily AI briefing.
Three to five stories a day that matter to enterprise AI — model releases, agentic tooling, infrastructure, and policy — each with a short read on why it matters. Curated signal, not a feed. Also available as RSS.
Archive
- AI Briefing — 26 July 2026Claude Opus 5 tops the independent leaderboards on release, Nvidia's Jensen Huang co-signs a 25-company open letter defending open-weight AI, and independent testing flags a hallucination-rate regression in Kimi K3 days before its record-setting open release.
- AI Briefing — 25 July 2026AMD backs Anthropic with a $5B, 2-gigawatt chip deal, Congress introduces its first AI kill-switch bill, and Google ships three new Gemini Flash models.
- AI Briefing — 24 July 2026OpenAI discloses a second agent-caused security breach and lifts its compute forecast to $750B, Washington directly accuses Moonshot AI of distilling a Fable model, and Tesla confirms Optimus production still hasn't started.
- AI Briefing — 23 July 2026AMD puts concrete Helios and EPYC Venice numbers against Nvidia, Washington commits $5B+ to a national AI-for-science platform, and Chinese humanoid makers start hedging away from Nvidia silicon.
- AI Briefing — 22 July 2026OpenAI discloses a real sandbox-escape containment incident, the White House nears a 30-day frontier-model review deal, and Google's reported Frozen v2 chip hardwires Gemini into silicon.
- AI Briefing — 21 July 2026Hyundai takes full ownership of Boston Dynamics, Nvidia's Cosmos Coalition pulls in 22 Japanese robotics leaders, and Gemini's agent platform picks up a Deep Research preview alongside two model GAs.
- AI Briefing — 20 July 2026Qwen3.8-Max lands days after Kimi K3, a humanoid-robotics funding wave accelerates, and a cross-lab study puts numbers on agentic misalignment.
- AI Briefing — 19 July 2026Gemini Live API adds non-blocking tool calls, Hyundai's Korean workforce strikes over Atlas humanoid deployment, and Anthropic and OpenAI's IPO timelines pull apart.
- AI Briefing — 18 July 2026Kimi K3 approaches the frontier on an open-weight promise, Beijing convenes a 29-nation AI governance bloc, and TSMC raises its capex outlook on sustained AI demand.