Introducing computer use in Gemini 3.5 Flash
Matters most to teams building internal automations, since it lets an AI agent actually operate a browser, phone, or desktop app on someone's behalf (clicking, filling in forms, navigating screens) rather than just answering questions. The built-in adversarial-training safeguards are relevant to anyone evaluating whether such an agent is safe to deploy.
What changed
Source summary
Google made "computer use" a built-in capability of Gemini 3.5 Flash, letting developers build agents that see, reason, and act across browser, mobile, and desktop apps, with adversarial-training safeguards.
Why it matters to your work
Practical interpretation
Matters most to teams building internal automations, since it lets an AI agent actually operate a browser, phone, or desktop app on someone's behalf (clicking, filling in forms, navigating screens) rather than just answering questions. The built-in adversarial-training safeguards are relevant to anyone evaluating whether such an agent is safe to deploy.
Source & availability
This is a summary of a published provider announcement, with practical context from SCALA’s feed.
Plan, region, workspace and rollout conditions may affect access in Australia. Check the original source for current availability.
blog.googleThis update is primarily for technical teams.