Brief IA

Chrome for Android: Gemini and Auto Browse Launch with Limited Access for US Subscribers

🛠️ AI Tools·Tom Levy·

Chrome for Android: Gemini and Auto Browse Launch with Limited Access for US Subscribers

Chrome for Android: Gemini and Auto Browse Launch with Limited Access for US Subscribers
Key Takeaways
1Auto browse requires explicit permission for sensitive actions, and Google says it is training its models against query injections.
2The feature is only available on Android in the United States with an AI Pro or AI Ultra subscription, with no timeline for other regions.
3On Android, Gemini in Chrome summarizes pages, answers questions, connects to Calendar/Keep, and generates images via Nano Banana.
💡Why it mattersGemini is arriving on mobile Chrome with announced agentic capabilities and safeguards, but auto browse remains limited to U.S. subscribers, with no date for other countries.
Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄
Full Analysis

Gemini is now accessible in Chrome on Android, following its desktop launch. Google highlights a suite of assistance tools and an agentic auto browse mode, which requires validations for sensitive actions and is only available in the United States with AI Pro or AI Ultra subscriptions. The company claims to have trained its models against threats such as query injections and does not provide a timeline for availability outside the United States.

Google trains its models against injections and requires validations

Auto browse seeks explicit permission before carrying out operations deemed sensitive, such as during a payment. Google asserts that it designed the integration of Gemini in Chrome with security and protection in mind and states that it has trained its models to recognize common threats, such as query injections. These attacks against AI chatbots can lead to leaks of sensitive data and other security risks. A recent case illustrates the potential misuses of agents: an agent hacked the website of an Australian gym to secure a spot in a class. The use of these systems therefore requires caution, as unpredictable or unintentional behaviors may occur.

Auto browse is only offered to AI Pro/Ultra subscribers in the United States

Auto browse is available to all Android users located in the United States, provided they have an AI Pro or AI Ultra subscription. Google has not provided a deployment timeline for other countries. Agentic AI on smartphones is described as increasingly common, with the latest Google Pixel 11 models cited as examples thanks to Gemini Intelligence.

Gemini in Chrome summarizes, answers, connects, and generates visuals

On Android, Gemini in Chrome can summarize lengthy pages and respond to targeted questions. It connects to Google services like Calendar and Keep without leaving the current tab. For creation, it can generate images using the Nano Banana generator; Google notably provides the example of an infographic produced from a page reviewed for an exam. The agentic auto browse feature offers the ability to delegate certain tasks, such as booking tickets for events or updating online orders. Google announced that Gemini would be available in Chrome for Android, following its initial appearance in May during its I/O conference. The mobile availability comes after a launch on desktop Chrome earlier this year.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.