Google is making changes to its AI Overviews feature after receiving "odd and erroneous" responses. AI Overviews, introduced last month, display AI-generated text at the top of search results with links to external websites. These summaries are generated using Google's large language model (LLM), Gemini, which aims to provide a quick overview and links for further information. However, concerns have been raised about the accuracy of the information and the difficulty of verifying it. Despite criticisms, Google claims that AI Overviews can enhance the search experience, although it acknowledges occasional inaccuracies.

In response to the feedback, Google has made several technical improvements, including limiting the inclusion of user-generated content and refining the triggering of AI Overviews. There are also concerns about the impact on websites' ad revenue since AI Overviews may reduce the number of visitors to their sites. Google, however, claims that the links included in AI Overviews receive more clicks. These changes come amidst increasing competition from other companies, such as OpenAI, that are developing similar generative AI technologies.