Gemini can now automate some multi-step tasks on Android

Read the full articleGemini can now automate some multi-step tasks on Android on TechCrunch

What Happened

Gemini on Android will be able to automate tasks involving rideshare requests, or grocery or food delivery, says Google.

Our Take

Finally, something actually useful. Not another chatbot—a thing that calls Uber for you. That's where AI wins: boring, repetitive multistep stuff. The catch? It has to work flawlessly or trust breaks instantly. Failed automation is worse than no automation. Google's testing with delivery and rideshare because they fail gracefully. The moment Gemini automates finances or health decisions, brittleness shows up hard.

What To Do

Don't build automation features until you prove 99%+ reliability in staging.

Builder's Brief

Who

mobile app developers handling transactional flows

What changes

Apps must expose structured intents or risk being bypassed by agent-layer orchestration

When

months

Watch for

Google publishing an official Android Agentic API or Intent schema for third-party apps

What Skeptics Say

Google has previewed Android agentic features repeatedly without broad rollout; real-money actions like rideshare and food orders have extremely low tolerance for agent errors that Google has not demonstrated it can meet.

Cited By

TechCrunch Gemini can now automate some multi-step tasks on Android

React

Newsletter

Get the weekly AI digest

The stories that matter, with a builder's perspective. Every Thursday.

Loading comments...