IBM Introduces Identity and Control for AI Agents, Securing Frontier Intelligence for the Enterprise

What happened

At TechXchange 2026, IBM announced new capabilities to help developers, IT and security teams deploy AI agents with clear authority and accountability, combining agent discovery, verified identities, access controls and performance evaluation so organizations can see agents, control their access and trace their actions.

Confirmed

Global impact / market context

Enterprises are deploying AI agents faster than they can verify them, and persistent credentials and unclear ownership make them hard to govern. IBM's tools aim to give each agent a governed identity and enforce least-privilege access, which means limiting each agent to only the data it truly needs, with real-time revocation.

Analyst inference

Gartner expects the average Fortune 500 enterprise to use over 150,000 AI agents by 2028, yet only 13% of organizations believe they have the right AI agent governance in place. That gap suggests rising demand for controls like those IBM is announcing, across Amazon Bedrock, Microsoft Foundry and Google Gemini.

Analyst inference

What to watch

1. IBM watsonx Orchestrate is designed to unify discovery, containment and evaluation across Amazon Bedrock, Microsoft Foundry and Google Gemini, letting enterprises see, test and account for agents across those platforms. Confirmed
2. IBM proposes extending HashiCorp Vault's established secrets management to AI agents, giving each agent a governed identity and enabling real-time revocation when an agent attempts an unauthorized action. Proposed
3. Watch whether enterprises adopt these controls as agent counts grow; if governance remains weak, unmanaged agents could reach sensitive systems, raising security and compliance costs and shaping spending on identity and access tools. Analyst inference

Evidence

• Original source link unavailable