Previously a standalone experimental model, computer use is now a built-in tool in production Gemini 3.5 Flash, letting developers deploy agents that control web, mobile, and desktop via the Gemini API and Enterprise Agent Platform.
Why it matters: Rolling computer use into a fast, low-cost production API significantly lowers the barrier for building browser-based marketing automation agents.
Designers can now embed live React code directly in the Figma canvas, add motion without leaving the tool, and generate custom AI plugins from natural language, collapsing three separate design-to-code handoff steps into one shared environment.
Why it matters: For teams building web products, this tightens the design-to-code cycle and gives AI agents richer design-intent context, reducing revision friction.
The new app (testing with select creators) brings Facebook's AI assistant into a dedicated interface for content ideation, audience growth, and performance analysis — Meta's move to prevent creators from reaching for ChatGPT as a third-party layer.
Why it matters: Native AI in Facebook's creator layer reshapes how brands brief and manage creator partnerships — relevant context for marketers running influencer or paid social programs.