Tips on how to Govern AI Brokers Earlier than They Go Rogue

Synthetic intelligence brokers are being deployed quicker than enterprises can safe them, creating blind spots throughout entry, possession and governance. An id safety material constructed on visibility, management and governance is the reply, mentioned Arkadiusz Krowczynski, principal product acceleration specialist at Okta.

See Additionally: How Unstructured Knowledge Chaos Undermines AI Success

Identification material spans three layers: visibility into the place brokers are and who owns them; management over what functions and information they’ll entry; and governance to maintain that entry safe over time, with entry critiques and a shutdown mechanism if an agent goes rogue.

With that material, “you’ll be able to safe end-to-end menace for AI brokers earlier than, throughout and after authentication. You’ll keep safe, however you’d nonetheless have the ability to innovate, work quicker and undertake capabilities of AI brokers,” Krowczynski mentioned.

On this video interview with Info Safety Media Group on the Gartner Identification and Entry Administration Summit, Krowczynski additionally mentioned:

• Why the id battlefield has moved past phishing to AI-weaponized token-based entry;
• Governance controls for AI brokers throughout third-party suppliers and provide chains;
• Dangers from enterprise groups deploying AI brokers with out IT visibility.

Krowczynski manages the high-stakes intersection of engineering, product technique and World 2000 C-suite necessities. He’s targeted on translating advanced technical highway maps into tangible enterprise outcomes, de-risking digital transformation and accelerating market velocity for enterprise-scale id and safety.