SBOMs Alone Can't Reveal Supply Chain Attack Scope
A real-world incident at a large-scale SaaS environment exposed critical gaps in supply chain attack response. The Shai-Hulud worm arrived in a package displaying all hallmarks of legitimacy: more than 3,400 GitHub stars, a credible-looking repository, and a valid SHA hash. The danger lay hidden three layers deep—a malicious transitive dependency that didn’t surface in standard dependency audits since it wasn’t a direct import. When a behavioral scanner detected suspicious activity, incident responders initially dismissed the alert; an SBOM reported the environment as clean. Later analysis revealed the SBOM was two development sprints out of date, providing no protection against an active attack.
Without real-time visibility into executing code, responders couldn’t identify which services were running the compromised package or determine the true scope of compromise. The team manually correlated logs across more than 400 services to piece together a picture. Tracing the attacker proved fruitless—the repository had been copied from legitimate code, and the publisher’s identity was entirely fabricated.
Deepam Kanjani of Atlassian, presenting the case study at ISC2 Strategic Congress 2026, concluded that SBOMs and dependency inventories provide insufficient visibility for incident response. Organizations need runtime monitoring to see what code actually executes in production. His recommendations: maintain deployed SBOMs within 72 hours of current state, detect unexpected module loads within 90 seconds, and flag lookalike package names within 24 hours of publication. The critical insight—incident response must focus on the runtime execution graph, not static dependency files.
Read the full article
Continue reading at Dark Reading →This is an AI-generated summary. Read the original for the full story.