Agentic development is pushing more code through CI than ever, and platform teams are the ones calling for SHA pinning across their organizations. GitHub supports it as an org-level policy, hardening guides recommend it, and after the tj-actions and the Trivy compromise, few teams argue that it isn’t a best practice.
But SHA pinning solves a narrower problem than most teams think. Pinning guarantees that the bytes don't change, not who wrote them, or whether they were part of the project you think you're running. Since any fork commit is reachable from the upstream repository, a pinned SHA can be an attacker's commit, and it'll look correct in your review, audit log, and policy scanner. Even a Docker-based action, when pinned to a SHA, still builds its image at runtime.
In this webinar, we’ll draw the line between immutability and provenance, evaluate GitHub's 2026 dependency-locking roadmap, and weigh the alternative approaches to trust.
We'll discuss: