Indian firms advancing AI adoption but lagging in identity governance, finds Okta survey

Okta’s latest research shows Indian organisations are embracing AI-powered digital experiences, but many still lack the governance, visibility and identity security capabilities required to securely manage AI agents and non-human identities as enterprise AI adoption accelerates.

Indian organisations are making significant progress in preparing for AI-powered digital experiences, but many are yet to establish the governance, visibility and access controls needed to securely manage AI agents and non-human identities, according to new research released by Okta Inc. The findings suggest that while enterprises are accelerating AI adoption across business functions, governance capabilities are still evolving to keep pace with deployment.

The research, based on polls conducted during Okta's AI Identity Summits held in Mumbai and Bengaluru, found that 73% of respondents said their organisations are either very prepared or somewhat prepared to deliver AI-powered digital experiences across applications, customer services and AI agents. However, the survey also points to a widening gap between AI adoption ambitions and the security and governance frameworks required to support trusted AI implementation. More than 120 technology and security professionals participated in the polling exercise.

Commenting on the findings, Matthew Graham, Chief Security Officer, APJ at Okta, said, “As organisations move beyond experimentation and begin embedding AI into business processes, governance becomes just as important as innovation. The organisations that will realise the greatest value from AI will be those that build visibility, accountability and security into their AI initiatives from the outset.”

Visibility and control over AI identities remain limited

Despite growing confidence in AI adoption, the survey found that many organisations continue to face challenges in monitoring and controlling AI agents operating across their IT environments. Only 14% of respondents said they have complete visibility into all AI agents and non-human identities within their organisations.

The research also revealed that only 17% of respondents are able to restrict or suspend an AI agent across all environments if required, while 51% said they could perform such actions only in some cases. According to Okta, these findings indicate that governance capabilities have not yet matured at the same pace as AI deployment, particularly as enterprises integrate AI into customer engagement, internal operations and business processes.

Confidence in identity governance still developing

The survey highlights similar trends in the management of non-human identities, including AI agents, service accounts and application programming interfaces (APIs). Only 17% of respondents said they were very confident in their organisation's ability to manage and secure these identities, while 53% reported being somewhat confident.

Confidence in AI access governance was also relatively low. Just 14% of respondents said they were very confident that AI agents are granted access only to the systems and data necessary to perform their designated tasks. The findings suggest that organisations are still strengthening policies and controls around identity governance as AI deployments become more widespread.

According to Okta, ensuring appropriate access management and visibility will become increasingly important as AI agents take on greater responsibilities within enterprise environments.

Identity security becomes central to enterprise AI

Reflecting on the changing priorities of organisations adopting artificial intelligence, Graham added, “The conversation around AI is shifting. The focus is no longer solely on adoption. Organisations are increasingly asking how they can govern AI consistently, manage access appropriately and maintain trust as AI agents become more embedded in everyday operations. Identity has an important role to play in answering those questions.”

Okta said the survey findings reinforce the growing role of identity security in enabling responsible AI adoption. As AI agents, APIs, service accounts and other non-human identities become increasingly connected to enterprise systems and sensitive data, organisations will require stronger lifecycle management, greater visibility and more robust access governance to minimise security risks while maintaining compliance.

The company's AI Identity Summits in Mumbai and Bengaluru brought together technology and cybersecurity leaders from across India to discuss identity security, AI governance and strategies for building trusted foundations for enterprise AI adoption. According to Okta, the research reflects the growing recognition that effective identity management will be a critical component of scaling AI securely across organisations.