Google is reducing the use of AI-generated answers in search results following instances of incorrect information, prompting technical changes to improve reliability.
Google is scaling back the use of AI-generated answers in its search results after several high-profile errors. This decision follows incidents where the AI provided incorrect information, including telling users to put glue on their pizza and incorrectly stating that Barack Obama was Muslim. The AI answers, also known as “AI Overviews,” had been integrated into search results in the United States for two weeks before the reduction was noticed.
Google’s head of search, Liz Reid, confirmed in a blog post that the company is cutting down the AI Overviews, especially those using social media posts as sources and those related to health topics. This move includes over a dozen technical changes meant to improve the system’s reliability.
The tech giant’s initial rollout of AI products has faced challenges, similar to previous issues like the temporary block on its AI image tool following accusations of bias. Despite these setbacks, Google announced over 100 AI-related features at its I/O conference, signaling its ongoing commitment to integrate AI into its services.
