Brakes built for speed: Why least privilege lets monday.com’s AI teams move faster

Permission management is emerging as a crucial safeguard as artificial intelligence agents proliferate across enterprise workflows, according to Ryan Nigro, security trust and enablement team lead at monday.com. These AI agents operate with the permissions granted to them by humans, rendering traditional permission management challenges even more pressing in a machine-speed context.

Nigro argued that strict governance protocols are not impediments to rapid progress but rather essential safeguards. He likened automobile brakes to their function: not to impede forward motion but to ensure safe acceleration. During a recent interview on theCUBE, SiliconANGLE Media's livestreaming platform, Nigro elaborated on the concept of least privilege, a cybersecurity principle initially conceived for human users.

As AI agents perform tasks akin to human employees, misconfigured permissions could pose significant risks, Nigro explained. He recommended a phased rollout of AI tools, each granted only the minimum access necessary, expanding permissions incrementally as the results validate the approach. Unplanned personnel changes can exacerbate permission issues, as employees moving between departments may retain outdated access rights within organizations that lack rigorous permission management systems.

IBM's research indicates that 80% to 90% of security breaches result from human error due to over-permissioning. Nigro emphasized that proper permissions should grant users the bare minimum required to perform their roles without enabling them to jeopardize the system’s integrity. Nigro’s insights were part of Monday.com's Elevate event coverage, sponsored by the company on theCUBE.