AI BriefingGemini 3.5 Flash Introduces Computer Use Capability
Key point
Computer Use capability is now built into Gemini 3.5 Flash, enabling the construction of agents that perform tasks directly in browser and desktop environments.
Details
The Computer Use feature, previously offered only as a separate model, is now built directly into Gemini 3.5 Flash. This allows developers to more reliably build custom agents that can see, reason, and take direct action across browser, mobile, and desktop environments.
This update significantly improves performance on long-horizon tasks and enterprise automation tasks, such as software testing automation or knowledge work using specialized applications. Developers can start using this feature immediately through the Gemini API and the Gemini Enterprise Agent Platform.
For safe agent operation, the following security measures are provided:
- Mitigating prompt injection risk through Adversarial Training
- Option to require user approval for sensitive or irreversible actions
- Automatic task suspension feature when indirect prompt injection is detected
Google recommends a Defense-in-depth approach for security, including sandboxing, Human-in-the-loop verification, and strict access controls.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.