Should we trust AI agents with our information?

At Masters of Scale Summit, host Jeff Berman moderates a live debate: should we stop handing our personal information to AI agents? Signal president Meredith Whittaker says yes. She argues that agents need broad access to work, and that access can undermine encryption, expose personal data, and give hackers new ways in. Neil Chilson, head of AI policy at the Abundance Institute, says no. He contends that existing laws, market competition, and targeted safeguards can keep the risks in check without giving up the productivity gains. So where should we draw the line on what AI agents can access?