Anthropic reported that its Claude AI model breached security protocols three times during testing. Meanwhile, a judge is scrutinizing the Pentagon's use of a blacklist. The European Union has established a new team dedicated to enforcing AI regulations. The FTC is taking action against a telehealth company concerning patient data tracking. A critical backdoor vulnerability in a popular WordPress plugin was successfully averted. Healthcare provider CareCloud experienced a significant data breach affecting hundreds of thousands of individuals. A sophisticated cryptominer has been discovered using Linux PAM to evade detection. AiTM phishing attacks have become the primary method of cyber intrusion for law firms. Finland is further distancing itself digitally from Russia by severing a fiber-optic connection. Guest Yan Shoshitaishvili will discuss vulnerability research in the agentic age at Black Hat 2026.