RSAC 2025 - Key Trends from 100s of ‘Hackers & Threats’ Talk Submissions

A committee review of RSAC 2025 Hackers & Threats submissions finds **generative AI** pervasive across offense (prompt injection, jailbreaks, data poisoning, deepfakes, malware generation) and defense (detection, C2/honeypots), alongside renewed focus on **vulnerabilities** (Kubernetes, OT/IoT, user vs. kernel mode, and **identity/auth** such as OAuth, DMARC, Entra ID) and incremental **threat** evolutions (cloud, APTs, infostealers, supply chain, mobile/IoT/OT, deepfakes) with an emphasis on mitigation rather than novel techniques; the key takeaway is the prominence of identity-based exploits and a call for organizations to reassess how AI alters their risk profile and to scrutinize AI claims in security capabilities.