Google has started testing a new AI-only search tool, as part of its plan to rapidly increase the use of artificial intelligence across its platforms.
Powered by Google’s latest Gemini 2.0 agent, the new “AI Mode” works by providing information to users based on what is available on the internet, rather than showing them a list of links. Instead of the traditional list of links, this new feature relies on Google’s massive search index to generate answers to queries in a conversational style similar to chatbots like ChatGPT.
“You can ask anything on your mind and get a helpful AI-powered response with the ability to go further with follow-up questions and helpful web links,” the search giant announced in a blog post on Wednesday.
“AI Mode is particularly helpful for questions that need further exploration, comparisons, and reasoning. You can ask nuanced questions that might have previously taken multiple searches – like exploring a new concept or comparing detailed options – and get a helpful AI-powered response with links to learn more.”
AI Mode is currently available only to users who pay for Google One AI Premium. However, it is likely to follow a similar launch pattern to previous AI tools, which become more widely available if successful.
Google also unveiled new updates to its AI Overviews, which appear at the top of the search results page during normal web searches. The overviews will now appear more frequently and will cover areas like mathematics and coding. Users will no longer need to be signed in to see the AI Overviews. These changes will impact the roughly 5 billion people globally who use Google, regardless of whether they have an account.
Google’s AI features have faced criticism from publishers and other websites, as they are designed to keep users on its own platform through so-called zero-click searches. The online education firm Chegg recently filed a lawsuit against Google, claiming that its referral traffic was being damaged by AI Overviews. These AI tools have also been criticized for providing incorrect information, such as suggesting that users eat rocks and stick cheese to pizza with glue. Google stated that these examples, which were widely shared on social media last year, were “not representative of most people’s experiences.”
