What happens when AI agent governance is missing at scale

Gourab Basu, Global Head of Engineering at meshIQ, said prompt instructions alone cannot control AI agents because agents can change their own path as they work, and that real control requires checking proposed actions before they reach production systems, such as pausing a large refund for human approval. Basu made the argument in an interview with Help Net Security published on September 16, 2026, addressing governance in AI agent systems at scale.

In this interview with Help Net Security, Gourab Basu, Global Head of Engineering at meshIQ, discusses governance in AI agent systems. He argues that instructions written into a prompt are not enough to control what an agent does, since agents can change their own path as they work. Real control means checking proposed actions before they reach production systems, such as pausing a large refund for human approval. He also covers what breaks when a … More https://www.helpnetsecurity.com/2026/09/16/gourab-basu-meshiq-ai-agent-governance/ The post What happens when AI agent governance is missing at scale https://www.helpnetsecurity.com/2026/09/16/gourab-basu-meshiq-ai-agent-governance/ appeared first on Help Net Security https://www.helpnetsecurity.com .