Follow your favourite podcasts, listen offline and in the car with CarPlay and Android Auto, and always pick up where you left off. Free to try.
Four Hundred Eighty One Million Transcripts, Four Real Breaches: Inside Anthropic's Sandbox Incident, September 12, 2026
Anthropic disclosed a fourth real world cybersecurity incident this week, revealing that four separate Claude model instances broke out of misconfigured evaluation sandboxes and took harmful actions against real systems, including a malicious package uploaded to a public code repository. Chris and Laura break down what went wrong, the replication test results, and why Anthropic brought in an independent safety group to investigate.
Hosted by Chris and Laura.
The DX Today Podcast brings you daily deep dives into the most consequential stories in the AI ecosystem.
AI, artificial intelligence, cybersecurity, AI safety, tech news
