Agentic AI , Artificial Intelligence & Machine Learning , Events Rethinking Cybersecurity With AI Agents Anthropic's Jason Clinton Discusses the Benefits and Challenges of AI Agents Jason Clinton, CISO, Anthropic AI agents will be crucial in the software development life cycle to eliminate bugs, improving the quality of software, which could significantly reduce security vulnerabilities. Although managing AI agents for identity and access controls will be hard, said Jason Clinton, CISO at Anthropic. See Also: How Generative AI Enables Solo Cybercriminals Despite the concerns, AI agents can be deployed to secure operational technology networks and assist with remediation tasks, ultimately speeding up response times to vulnerabilities, Clinton said.

"As defenders, we need to be thinking about what we can do to adopt models on the defensive side to help protect our networks and help protect our systems against that. So, there's lots of things that we can do in that space that sort of help," Clinton said. In this video interview with Information Security Media Group at RSAC Conference 2025, Clinton also discussed: The future of software development; The impact of AI on offensive cyber capabilities; Virtual employees and identity management.

At Anthropic, Clinton leads efforts to secure cutting-edge AI infrastructure. Before joining Anthropic, he headed Chrome infrastructure security at Google, playing a key role in safeguarding the browser's backend systems. His earlier work includes enhancing the security of ChromeOS build processes, fortifying Android Pay, and developing secure digital coupon systems.

He also worked on Beowulf cluster computing and secure thin client architectures, reflecting a long-standing focus on system integrity and scalability.