Commvault has launched a new cyber resilience program designed to help security teams prepare for the growing use of artificial intelligence by cybercriminals.
Known as Minutes to Recovery™, the live simulation places participants inside an AI-driven cyberattack, requiring them to attack, defend and ultimately recover systems during a realistic two-hour exercise.
The scenario begins with participants using frontier AI tools similar to those employed by threat actors to create phishing campaigns and initiate an attack. Teams then move into defensive roles, making rapid detection and response decisions before undertaking the recovery process to restore systems while ensuring threats are not reintroduced.
The exercise has been developed to reflect the changing nature of cyber threats, where the time between vulnerability disclosure and exploitation continues to decrease. Commvault says this window has reduced to just 29 minutes during 2025, placing greater emphasis on operational readiness and recovery capabilities.
Beyond technical skills, the simulation also examines how effectively security, IT and business teams collaborate during an incident. Participants are challenged to make decisions with incomplete information, exposing potential weaknesses in communication, incident response and recovery planning.
At the conclusion of the exercise, organisations receive a Mean Time to Clean Recovery (MTCR) benchmark that measures their ability to restore verified clean systems under pressure.
Anna Griffin, Chief Market Officer at Commvault, said organisations need to demonstrate—not simply assume—that recovery plans will work during an actual incident.
“The question organisations need to answer is no longer, ‘Do we have a recovery plan?’ Instead, they should be asking, ‘Can we prove it will work under pressure?'”
The program is available globally in six languages and will also be delivered through Commvault’s partner ecosystem.
Kyndryl Vice President of Security and Resiliency Allen Downs said the experience helps organisations validate resilience against increasingly sophisticated AI-enabled attacks.
