Anthropic Wants You to Know Its New AI Model Is Definitely Not Too Dangerous to Release by Webb Wright from Gizmodo on 2026-06-30 21:00 (#76P01) Claude Sonnet 5 delivers impressive agentic capabilities at a relatively low cost. It's also really bad at cybersecurity-probably for the reason you'd expect.
The Pentagon’s Case Against Anthropic Isn’t Going Well by AJ Dellinger from Gizmodo on 2026-07-30 21:30 (#77BSG) Shocking this administration isn't a reliable narrator.
