Keepit, the only vendor-agnostic cloud dedicated to SaaS data protection, announced the results of its “Peer Insights on the AI Adoption and Disaster Recovery Gap” study. SaaS data protection is considered a top priority when implementing AI solutions, but a survey of senior IT decision makers reveals a gap between perceived readiness and tested and validated disaster recovery capabilities.
Also read: AiThority Interview with Glenn Jocher, Ultralytics Founder and CEO
Key findings:
- 94% of respondents said they are confident that their current disaster recovery plan covers scenarios involving agent AI systems.
- 33% of IT and security leaders say they have only partial control over the use of agent AI in their organizations.
- 56% of respondents consider SaaS data protection and disaster recovery a top priority when implementing an AI solution.
- Only 41% of respondents have significantly changed their approach to disaster recovery planning due to accelerated adoption of AI.
- 32% of respondents test their disaster recovery plans monthly.
According to the survey, 94% of respondents say they are confident that their disaster recovery plans cover agent AI systems, but only 32% test those plans monthly. This gap between trust and verification raises concerns about an organization’s ability to actually recover when failures occur, especially as AI-driven automation makes systems more interdependent and accelerates the spread of errors.
Governance and testing lag behind AI adoption
While the introduction of AI raises the bar for governance and recovery planning, research shows that many organizations have not evolved their disaster recovery approaches accordingly.
The survey found that 33% of IT and security leaders say they have only partial control over the use of agent AI in their organizations, and 52% doubt whether their company’s recovery plans include agent AI. scenario. Only 41% of respondents said their approach to disaster recovery has changed significantly as a result of implementing AI.
“Organizations must place increased emphasis on creating long-term, structured, and tested disaster recovery plans,” said Kim Larsen, group chief information security officer at Keepit. “It also means focusing on governance and accountability,” he continued. “At the end of the day, it’s important to know who will be responsible for recovery and which systems will be restored first if multiple systems are affected. Delaying decisions will make recovery take longer than necessary.”
Build confidence in your recovery
Confidence in your recovery is built through regular testing. This study showed that there is a gap between reliability and tested recovery ability. Although backups are common, recovery capabilities are not consistently understood, tested, or verified. These findings are supported by the Keepit Annual Data Report 2026, which showed that recovery practices remain a work in progress for many organizations, especially smaller ones.
“One of the challenges we face in deploying agentic AI is properly securing identity and access management. The Keeit Annual Data Report 2026 found that identity system recovery was tested one-fourth as frequently as productivity system recovery, highlighting a lack of recovery maturity. This is of particular concern for identity applications that are critical to managing agentic AI. Securing them is paramount, as access to applications can be cut off and operations halted,” adds Kim Larsen.
Also read: The infrastructure war behind the AI boom
[To share your insights with us, please write to psen@itechseries.com]
