Catching rogue AI behavior with identity-aware analytics

When you look at your AI bill, it can be hard to tell if anything is amiss — whether an agent has gone rogue, an employee's usage has spiked tenfold, or something's shifted you didn't expect. According to Cloudflare, fifty-nine percent of organizations say knowledge gaps are their biggest obstacle to responsible AI governance. The company is launching two new capabilities to address this. First, Identity-aware AI Gateway ties every AI request to a verified employee through existing company identity systems like Okta, replacing shared API keys that obscure who spent what. Second, User Insights learns each person's and agent's normal behavior, then flags sessions that deviate sharply from their baseline — catching the spike before the invoice lands. Flexport, an early adopter, found that shared API keys made it nearly impossible to enforce access rules or see who was spending; adding identity controls solved both problems.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton