India, Nov. 11 -- As enterprises move beyond experimenting with artificial intelligence (AI) to deploying autonomous agents across business operations, a new cybersecurity challenge is emerging. Unlike conventional software, AI agents can interact with enterprise systems, modify configurations, access data, and execute actions with limited human intervention. The consequences of an error or a compromised agent could therefore unfold faster than organisations can respond.

Sanjay Mirchandani, President and CEO of Commvault, believes this is creating a new category of enterprise exposure that organisations need to recognise and address. He calls it "AI vulnerability", arguing that the growing autonomy of AI systems is fundamentally changing...