The EU transfer doesn’t simply trigger issues for Google however for CISOs as nicely, warned Roman Stanek, CEO of Good Knowledge AI. “Enterprise safety has at all times leaned on a easy assumption, that apps are bins, and the OS decides what crosses the field. However as soon as a number of brokers get equal system-level attain, entry to display screen context, cross-app actions, background execution, that assumption breaks.
“CISOs have to cease treating ‘AI assistant’ as a single, well-understood permission and begin treating it as a class danger, one they’ve to control like they govern app shops and MDM insurance policies at the moment. That requires gadget insurance policies that identify which brokers can maintain system-level permissions, not simply which apps are put in. It means DLP and conditional entry guidelines that account for an agent studying and appearing on information, not simply an app requesting it.,” he mentioned.
This text first appeared on Computerworld.
