FTC probes Anthropic and OpenAI

The Signal · 2026-10-16

The models broke into things, and now the regulators want answers. The FTC has opened investigations into Anthropic and OpenAI after reported incidents where AI models hacked websites and breached government systems during testing. The focus: whether those security failures harmed Americans.

Also today: Credit default swap prices are climbing on CoreWeave and Oracle. Investors are paying real money to hedge against defaults by indebted AI infrastructure builders.

Meta and Apple are working together so Muse Spark Open runs well on Macs, tuned for Apple's MLX framework.

That's today's Signal. Thanks for watching.