Google Launches Gemini Spark, Agentic AI Agent Taking Actions Without User Prompts

Google unveiled Gemini Spark at I/O 2026, a personal AI agent powered by Gemini 3.5 that autonomously performs tasks across Google Docs, Slides, and other applications without waiting for explicit user commands. The tool runs on dedicated Google Cloud virtual machines and is built on Google's Antigravity IDE, representing a shift toward agentic AI that operates more independently.

Why it matters: Agentic AI represents the next evolution in AI assistants—moving from reactive question-answering to proactive task execution—which could fundamentally change how knowledge workers interact with productivity software.