Okta Research: How Multi-Vendor AI Stacks up Identity Debt

Enterprises have not been shy about adopting AI but new research shows they may be accumulating what is called ‘identity debt’.

This refers to the unresolved identity and access management (IAM) issues that may build up because of improper permissions, authentication and governance. 

These concerns materialised with Okta’s recent research, which has revealed that enterprises across Europe, the Middle East and Africa are facing growing levels of such ‘identity debt’ as AI adoption gathers pace. 

Okta’s research is based on patterns across more than 20,000 organisations worldwide.

It also tells us that businesses are embracing increasingly complex AI ecosystems rather than putting all of their eggs in the basket of a single provider. This creates fresh security, governance and compliance challenges.

The research further highlights how AI adoption has evolved much further beyond early experimentation, as businesses shift from standalone chatbot tools to AI-powered systems capable of handling longer-running tasks and more sophisticated workflows with minimal human intervention.

‘Death of single-vendor lock-in’

One key research finding was that enterprises are not keen on the idea of standardising on a single AI platform. 

They are stringing together a variety of AI-native technologies with solutions from established enterprise software providers creating more flexible technology stacks.

In June 2026, exclusive single-platform AI deployments declined by more than 1.3% compared to the previous month. This underlines the rapid shift towards multi-vendor environments and what Okta notes as the “death of single-vendor lock-in”.

Data also shows that specialist AI companies – or the “pure-play disruptor companies” including the likes of Anthropic, OpenAI and Cursor – continue to gain momentum, while well-established technology companies like Microsoft and Google retain significant enterprise market share through their established customer relationships and integrated software ecosystems.

Governance and open standards will shape the future of AI

While adopting multiple AI platforms offers greater flexibility and innovation, it also introduces new operational and security risks, particularly for organisations operating under strict governance and regulatory requirements across EMEA.

Managing different AI tools can result in misconfigured permissions, inconsistent security policies and fragmented oversight. These issues increase the likelihood of security breaches and compliance failures.

The research suggests that identity management is becoming just as important as AI adoption itself. 

“AI is not just hand built or custom agents. For the majority of users, it will be embedded in the apps they already use today,” notes Richard Wainwright, EMEA CTO at Okta. 

“Users will demand that they can combine resources from multiple platforms to deliver better business processes.

“This is especially true in EMEA. Organisations will need to be able to show ownership, traceability and chain of custody for AI decisions that impact end users.

“To meet that goal, businesses need tooling that can manage the AI lifecycle across end user devices, custom built agents and within SaaS apps and they need to embrace open standards to facilitate that. 

“No company is going to use AI from only one vendor, so an open, independent platform that can integrate is critical.”