Gemini 3.5 Flash now includes a built-in computer use feature, enhancing automation and testing tasks for developers. This integration allows for improved functionality in various environments and addresses prompt injection risks through training and enterprise safeguards.
The Gemini 3.5 Flash model has introduced a built-in computer use functionality, previously available only in the standalone Gemini 2.5 model. This integration allows developers to build advanced agents that can perform tasks reliably across different platforms such as browsers, mobile, and desktop environments. With enhanced capabilities for function calling and using built-in tools, Gemini 3.5 Flash is now better equipped for long-horizon and enterprise automation tasks.
Developers can leverage the new computer use capability in Gemini 3.5 Flash for various applications including continuous software testing and knowledge work. The model analyses the app's features and audits its own documentation for accessibility, thereby improving the overall utility and functionality of the Gemini platform.
To address potential prompt injection risks associated with agents in live environments, Gemini 3.5 Flash employs targeted adversarial training. Additionally, it provides two optional safeguards for enterprises: requiring user confirmation for critical actions and stopping tasks if indirect prompt injection is detected. These features are recommended to be used alongside secure sandboxing and strict access controls.
Initial customer feedback indicates value generated through the implementation of the new computer use features. Developers interested in utilizing these capabilities can access demo environments through Browserbase, as well as reference implementation and documentation via the Gemini API and the Gemini Enterprise Agent Platform.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
Gemini 3.5 Flash now includes a built-in computer use feature, enhancing automation and testing tasks for developers. This integration allows for improved functionality in various environments and addresses prompt injection risks through training and enterprise safeguards.