Introducing computer use in Gemini 3.5 Flash

2026-06-24 · Source: The Keyword · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Software Development & Engineering · Depth: Intermediate, medium

Summary

Google DeepMind has integrated "computer use" as a built-in tool within Gemini 3.5 Flash, enhancing its agentic capabilities for cross-platform interaction. This feature, previously a standalone Gemini 2.5 model, now allows developers to build custom agents that can perceive, reason, and act across browser, mobile, and desktop environments. This integration significantly improves performance for long-horizon and enterprise automation tasks, such as continuous software testing and knowledge work in professional applications. To address prompt injection risks in live environments, Google employs targeted adversarial training and offers two optional enterprise safeguard systems: requiring explicit user confirmation for sensitive actions and automatically stopping tasks upon detecting indirect prompt injections. Developers can access this functionality through the Gemini API and Gemini Enterprise Agent Platform, with a demo available via Browserbase.

Key takeaway

For AI Engineers developing automation solutions, Gemini 3.5 Flash's integrated "computer use" capability offers a robust platform for building agents. You can now create agents that interact across browser, mobile, and desktop environments, streamlining complex enterprise automation and software testing. Utilize the Gemini API or Enterprise Agent Platform, combining built-in safeguards with sandboxing and human-in-the-loop verification to manage prompt injection risks effectively.

Key insights

Gemini 3.5 Flash now natively supports "computer use" for building cross-platform automation agents with enhanced safety features.

Principles

• Agentic AI requires cross-platform interaction.
• Safety mitigation needs defense-in-depth.
• Enterprise automation benefits from integrated tools.

In practice

• Build agents for software testing.
• Automate knowledge work tasks.
• Audit documentation for accessibility.

Topics

• Gemini 3.5 Flash
• Computer Use
• AI Agents
• Enterprise Automation
• Prompt Injection
• AI Safety

Code references

• google-gemini/computer-use-preview

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Engineer, Automation Engineer, MLOps Engineer

Related on AIssential

• Introducing computer use in Gemini 3.5 Flash — Gemini 3.5 Flash now natively supports "computer use" for building advanced, cross-platform automation agents with enhanced safety features.
• Google bakes computer control directly into Gemini 3.5 Flash, letting the model see and operate your screen — Gemini 3.5 Flash now directly controls computers, enabling autonomous agent development.
• Gemini 3.5: frontier intelligence with action — Gemini 3.5 Flash delivers frontier agentic and coding performance at high speed and lower cost, enabling complex workflow automation.
• Google makes an interesting choice with its new agent building tool for enterprises — Google's Gemini Enterprise Agent Platform enables scalable AI agent development for technical and business users.
• TAI #206: Gemini 3.5 Flash Is Stronger, But Flash Is No Longer Cheap — Google's latest Gemini models offer strong capabilities but come with significantly increased costs and deployment challenges.
• Google launches Gemini 3.5 Flash as its default AI model — Google's Gemini 3.5 Flash offers high-speed, cost-effective AI for agentic tasks, becoming the default model.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by The Keyword.