WIRED: Claude Misuse Now Everywhere, Meta Missed ~350 AI Child Abuse Ads
nordicinst · x · 2026-09-12
WIRED's weekly security roundup (also the final edition after a decade) covers a string of AI misuse incidents:
- Meta failed to catch roughly 350 AI-generated child abuse ads, some using images of real children including a European royal family member. The San Francisco City Attorney's Office ordered Meta to stop allowing such ads, and lawmakers plan to investigate.
- Anthropic's Claude has been misused for everything from hacks to bioweapons.
- The US disrupted the internet's biggest black market; a Conti ransomware hacker was sentenced.
- Sweden/EU are pushing AI safety: tighter rules, training, and deterrence.
- Meta also launched a personal AI agent for booking flights or selling cars, emphasizing security—while facing a proposed class action over alleged illegal Facebook data harvesting.
More from Companies & People
- Codex workshop at Chiang Mai University yields 10+ student projects, BigOwl wins — paw_lean · 2026-09-12
- 30 people in Santiago built apps entirely from phones in OpenAI Codex workshop — paw_lean · 2026-09-12
- Cluely CEO Roy Lee Melts Down Over People 'Not Taking Action' — JasonBotterill · 2026-09-12
- Palantir Foundry and NVIDIA cuOpt Power NVIDIA's Own Supply Chain Allocation — Puzzled-Ad-6854 · 2026-09-12
- AI Whistleblower interview: OpenAI scandal, AI cults, Neuralink & our last chance — IgorGabrielan · 2026-09-12
- Commenter ties Google DeepMind's decline to researchers' extraordinary bargaining power — S_OhEigeartaigh · 2026-09-12