Snyk & Anthropic: Preparing for the Coming Wave of Autonomous Attacks

AI has handed attackers a capability most security programs were never built to face: finding architectural flaws at machine speed. The same AI is widening the target at the same time, shipping code faster than any team can review it. Attackers are now hitting every layer at once, unfixed legacy bugs, new code as written, and application agents running in production, while agentic development accelerates faster than security programs can track it.

In this recorded session, The Register’s James Hayes is joined by Manoj Nair, Chief Innovation Officer at Snyk, and Alon Krifcher, Head of Applied AI at Anthropic, for a direct look at what is actually changing and what a credible response looks like in operation.

Watch the recording to hear what Snyk and Anthropic are each seeing, and what to do about it. You will learn:

• How the threat is changing at the model layer, and what it looks like by the time it reaches production

• Why continuous attacks require continuous validation, and where model-layer hardening ends and your own environment takes over

• The four actions that close the gap: discover the full attack surface, remediate the backlog, validate what is actually exploitable, and prevent what comes next