Today, AI agents are doing real work—accessing databases, operating across browsers—and they’re doing so by potentially performing thousands of actions per minute. So who is really in control of what they can do? For the latest episode of CRN’s Security or Else! podcast series, 1Password CTO Nancy Wang spoke with CRN Senior Editor Kyle Alspach about this brave new world of AI agents—and why securing them requires rethinking identity, access and control. Wang explains why agents can’t simply be treated the same way as human users and what the history of cloud security’s rise can teach us about the next wave of AI security—while also discussing why the controls and guardrails around an AI model deserve more attention than they’re getting. She also speaks about how 1Password is approaching its use of Anthropic’s Claude Mythos cyber model and where MSPs can add value—as well as why she is not a fan of doom-and-gloom messaging on cybersecurity. Please like, subscribe and follow Security or Else! for more conversations about what’s hype and what’s real—and what channel partners should focus on next in security and AI.