IBM: Agentic AI Is Reshaping Governance, Controls and Trust
TechTarget, Wednesday, October 14th, 2026
At IBM Think on Tour EMEA, executives said scaling agents safely is an operating-model problem that needs governance by design.
At IBM's Think on Tour EMEA 2026 in London, IBM's Ana Paula Assis said 80% of executives expect AI to drive revenue but only 20% can explain how.
IBM research found three-quarters of EMEA executives expect agents to complete full business tasks and decide without humans by 2027, and 80% say an AI-related problem has already reached the boardroom.
She called safe scaling an operating-model rather than a technology problem.
Atos' Michael Nimhauser said fear of losing control extends to infrastructure, models and orchestration, so governance must be built in by design, while Celerity runs governance reviews alongside value assessments.