GOOGLE DETAILS SECURITY MEASURES FOR CHROME'S AGENTIC FEATURES

Summary from the TLDR InfoSec newsletter

Chrome's upcoming AI “agentic” features will act on users' behalf to complete tasks such as shopping or filling out forms, while also reducing security risks. It uses separate models to check that planned actions align with user goals and to vet URLs, limits what page data agents can read or modify, and asks for explicit permission before accessing sensitive sites, using passwords, or completing purchases.