Cynthia Lee Vice President APAC, Delinea
“AI helps attackers run phishing and harvest credentials faster, while deepfakes and synthetic identities make lures harder to spot. If an AI identity is compromised, attackers may use it to trigger workflows and take privileged actions at machine speed. Delinea’s research shows 34% of Indian organisations caught their latest AI scope violation as it happened, against 20% globally. Enterprises need AI to control what those identities can do, applying least privilege and Zero Trust to every agent.
Enterprises need to know what data an AI tool can reach and for how long. Our research found 99% of Indian organisations have a formal AI data access policy, yet 84% say an AI tool accessed sensitive data beyond its intended scope in the past year. Continuous authorisation closes that gap by controlling what each identity can access when an action is attempted. As the DPDP Act takes effect, enterprises will be asked to prove who authorised each access decision and what the agent did.
Delinea controls what every identity, from AI agents to employees, can do after access is granted. We authorise each action just in time and apply least privilege across on-premises and multi-cloud infrastructure. Only 47% of Indian organisations can always trace a sensitive AI access event back to its approver. Delinea improves traceability by linking every access decision to an approver, with session visibility and a defensible record.”
