Snyk reposted this
“Detection was never the bottleneck.” That is how J C. Jason Clinton , Anthropic’s Deputy CISO, framed AI security in our announcement today. He is right. The bottleneck has always been what happens after a finding lands. Getting to a verified fix, in the workflow where developers and agents are already working, before the next commit ships. That is what we built with Anthropic. Starting today, Claude is embedded inside the Snyk AI Security Platform and Evo. Claude’s reasoning powers both ends of the loop. Discovery gets sharper because Claude understands intent across files. Remediation gets faster because Claude generates higher-confidence fixes inside the same workflow developers and agents already use. The reality this partnership is built for: 65-70% of new enterprise code is now AI-generated. Every AI model deployed brings nearly three times as many additional software components. 82% of those tools come from third-party packages. The agents shipping all of this operate almost entirely outside traditional AppSec tooling. (From Snyk’s 2026 State of Agentic AI Adoption Report, drawn from 500+ enterprise Evo environments.) Every CISO I speak to is asking the same thing. Help us govern AI development at the speed it is shipping. Brendan Putek and the team at Relay Network became a design partner because agentic development was outpacing their AppSec program. They shifted security from a retroactive gate to an integrated part of code creation. Now every Snyk + Anthropic customer can do the same. AI risk starts in development. That is where it has to be governed. Snyk knows security. The agent knows the app. Together they close the loop. Thanks to Jason Clinton, the Claude team, and everyone at Anthropic who partnered with us on this. More details: https://lnkd.in/eAETkNpe