What Identity Means in the Age of Agentic AI: Insights from Think 2026
IBM, Friday, May 8th, 2026
IBM unveils security solutions for managing non-human identities and agentic AI at enterprise scale.
At Think 2026, IBM highlighted how agentic AI is reshaping enterprises but traditional IAM systems weren't built to manage the explosion of non-human identities (NHIs) and autonomous agents.
With 45-90 NHIs per human identity and 92% of organizations lacking confidence in legacy IAM tools, IBM introduced Vault 2.0 and a unified Verify/Vault solution to enforce identity, access, and policy continuously at runtime.
The company recommends five imperatives: proving identity and intent at every step, preventing standing privilege through short-lived access, controlling access at runtime, making autonomy accountable with full auditing, and unifying control across environments. Real-world examples from regulated industries like healthcare demonstrate the critical need for this runtime security model as organizations deploy agentic systems in production.