Pentagon Said Never Again, White House Said Come In — Anthropic Now Has Two Doors
Emil Michael declared the Pentagon will never again rely on a single AI provider; the White House is simultaneously opening a civilian path for Anthropic's Mythos.
Technology
Waymo Recalled 3,800 Vehicles After One Drove Into Flood Water It Had Already Flagged
NHTSA filed a recall May 12 after a Waymo robotaxi in San Antonio slowed to flag flood water on April 20, then drove into it anyway and was swept into a creek.
Vercel's Breach Was a Supply Chain Attack — and the Paper Had Its Phase Wrong
Vercel disclosed publicly on April 19-20, via its knowledge base and TechCrunch, that a Context.ai supply chain attack exposed customer environment variables.
Meta's AI Productivity Claims Come With a Surveillance Methodology Inside
Meta's Model Capability Initiative records employee mouse movements, clicks, keystrokes, and screenshots, with no opt-out, to train AI on human computer behavior.
Vercel Found a Second Breach Wave Unrelated to the Context AI Attack
Vercel's expanded investigation turned up a separate set of compromised accounts predating the April Context.ai supply chain attack.
Anthropic Launched Glasswing While the NSA Quietly Tested Mythos
Anthropic launched Project Glasswing — a corporate critical-software coalition — while the NSA separately tested Claude Mythos, the model the Pentagon has blacklisted.
Waymo Says the Vehicle Recognized the Flood — the Override Did Not Respond
The Waymo that drove into standing water had already slowed to detect the flood — then overrode itself and kept going anyway.