Новые границы ИИ в Chrome: Gemini, Nano Banana и революция персонального интеллекта Translation: New Frontiers of AI in Chrome: Gemini, Nano Banana, and the Revolution of Personal Intelligence

Google is incorporating new AI tools based on Gemini into the Chrome browser. The key innovations include the image generator Nano Banana, «Personal Intelligence,» and «Auto Preview.»

The AI assistant, which was introduced in September 2025 as a floating window, is now anchored in the sidebar. Users can ask questions regarding the website they are exploring or other open tabs.

A new feature allows for the analysis of multiple tabs as a cohesive unit. When users access different pages from the same site, the digital assistant perceives them not individually, but as a contextual group.

Previously, Gemini in Chrome was only accessible to Windows and macOS users. Following the update, it is now available for Chromebook Plus users as well.

The corporation leverages the recently launched «Personal Intelligence» feature, which connects to Gmail, Search, YouTube, and Google Photos accounts, enabling users to ask questions based on personal data.

This feature will roll out in the upcoming months, permitting users to inquire about various topics, such as family schedules or drafting emails without switching to Gmail.

The tool allows users to edit and modify images, using other images found online as references.

The AI agent can autonomously carry out tasks such as clicking links, making purchases, or searching for discount coupons. It will request user intervention in cases that involve handling sensitive data.

AI browsers are emerging as a trend in the artificial intelligence sector, yet they pose significant risks to users. In December 2025, OpenAI revealed existing vulnerabilities.

The company acknowledged that «prompt injection» attacks represent a challenge.

«Such vulnerabilities, along with online fraud and social engineering, are unlikely to ever be completely eliminated,» said OpenAI representatives.

Both Anthropic and Google share a similar stance, focusing on multi-layered protection and regular stress testing.

Simultaneously, Google introduced Agentic Vision, a feature that enables a more detailed examination of files in AI-agent mode.

The company explained that next-generation LLMs like Gemini typically analyze information about the surrounding world from a single static perspective. If they overlook fine details, such as a serial number on a microchip or a distant road sign, they tend to «make assumptions.»

Agentic Vision in Gemini 3 Flash transforms image understanding from a static action into an active process. The model functions like a human analyst:

As a result, recognition accuracy increases by 5-10%.

Agentic Vision introduces a «think, act, observe» cycle in image comprehension tasks:

Gemini 3 Flash is trained to magnify images when it detects fine details.

The beta version of Agentic Vision is available for free on Google AI Studio, Vertex AI, Gemini API for developers, and the Gemini chatbot in Thinking mode.

As a reminder, in December 2025, Google released the language model Gemini 3 Flash and made it the standard in the Gemini app and AI mode in the search engine.