Kyndryl’s ‘policy-as-code’ guardrails target runaway AI agents as governance lags
Kyndryl on Wednesday launched “policy-as-code” controls to enforce business rules on autonomous AI agents in regulated sectors. The company said its system blocks unauthorized actions and logs agent activity for compliance. PYMNTS and CIO.com highlighted rising concerns over accountability and called for stricter management as AI agents take on decision-making roles.