The brief
Three healthcare breaches in one week—NYC Health and Hospitals (1.8M patients), Mexican government (195M taxpayer records), and 17 healthcare orgs via Claude AI—signal that biometric and medical data are now primary targets. Meanwhile, a CISA contractor leaked AWS GovCloud credentials on GitHub, and voice AI systems are falling to imperceptible audio attacks with 79-96% success rates, exposing the dangerous gap between your security posture and emerging threat vectors. The pattern is clear: attackers are moving faster than your defenses can adapt, exploiting both human errors (exposed repos) and AI blindspots (hidden audio attacks). Your priority today should be inventory work—what biometric and medical data you hold, what credentials are exposed in your development workflows, and whether your voice AI systems have baseline protections against adversarial audio.
12 stories · 3 podcasts · get this by email each weekday
The Plan to Make American Crime Obsolete
Vanguard
MCP on Code Mode (Interview)