Google has integrated native computer use capabilities into Gemini 3.5 Flash, establishing the feature as a built-in tool within the model architecture. Previously accessible only as a standalone component in the Gemini 2.5 series, computer use now operates natively alongside Gemini's existing strengths in function calling and grounded utilities like Search and Maps. This integration enables developers to deploy custom agents capable of observing, reasoning, and executing actions across browser, mobile, and desktop environments. The upgrade targets high-complexity workflows, including long-horizon enterprise automation, continuous software testing, and knowledge work across professional applications. Demonstrations indicate the model can analyze application interfaces to categorize features and audit documentation for accessibility compliance. Security protocols for agentic operations include targeted adversarial training to mitigate prompt injection risks. Google is releasing two optional enterprise safeguard systems designed to require explicit user confirmation for sensitive or irreversible actions and to automatically halt tasks when indirect prompt injection is detected. The company recommends a defense-in-depth approach, advising developers to supplement these safeguards with secure sandboxing, human-in-the-loop verification, and strict access controls. Computer use in Gemini 3.5 Flash is available via the Gemini API and the Gemini Enterprise Agent Platform. A demo environment hosted by Browserbase offers immediate capability testing, with reference implementations and documentation provided for integration.