Google is expanding its AI Overviews feature to over 200 countries along with broadening support for more languages, the company announced during the keynote of its I/O 2025 developer conference held in California on Tuesday. With this update, the AI-powered search tool will now be available in Arabic, Chinese, Malay, Urdu, and more languages, enabling users to see text in the local languages of each of their respective countries and get a quick overview of the topic.
Expansion of AI Overviews
Google announced the expansion of AI Overviews in a blog post. AI Overviews will now be available in more than 200 countries and over 40 languages. It is adding support for Arabic, Chinese, Malay, and Urdu which join existing language options such as English, Hindi, Indonesian, Japanese, Portuguese, and Spanish.
As per the Mountain View-based tech giant, more users can leverage the AI-powered feature to ask questions, quickly find information, and explore relevant websites in Google Search results. It enables them to see snapshots containing a brief overview of the topic they’ve searched for, along with external links for more information. It leverages the Gemini large language model (LLM) to fulfill search queries while also taking advantage of its multi-step reasoning capabilities.
Notably, AI Overviews was first announced in public access at Google I/O 2024 and its availability was initially limited to the US via Search Labs. Since its public rollout and expansion to more regions in recent months, the feature has witnessed a 10 percent increase in usage of Google for specific queries which display AI Overviews. As per Google, the US and India have been some of its biggest markets so far.
In addition to expansion of AI Overviews, the company is also making changes to the LLM which powers the feature. Starting this week, a custom version of Gemini 2.5 will power AI Overviews in Google Search in the US. Google says it is its most intelligent AI model to date and will help tackle more complex queries and questions. However, the company still advises caution as AI responses may include mistakes.
