Are AI labs ignoring cybersecurity experts?

Cybersecurity professionals contend that major AI labs — including OpenAI and Anthropic — are pursuing AI safety regulation without meaningfully including security practitioners in the process, and are focused on the wrong threat vectors. The podcast episode argues that while labs advocate for rules around existential and societal AI risks, operational cybersecurity concerns such as AI agent vulnerabilities, adversarial exploitation of models, and post-quantum cryptography readiness are being sidelined. Security experts assert they are largely absent from the policy rooms where AI governance frameworks are being drafted, creating a gap between emerging AI regulation and practical enterprise security risk management. This disconnect has direct implications for CIOs and security leaders deploying AI systems who must manage threats that current regulatory proposals may not address.