Google is preparing for the upcoming US presidential election by rolling out safety measures to more of its generative AI products. The company had already announced that it would be restricting Gemini queries related to the election, but is now applying additional restrictions to its other generative AI products.
These safeguards apply to search AI profiles, YouTube AI-generated profiles in live chat, Gems, and image generation in Gemini. As part of the restrictions, these AI products will not respond to election-related topics.
Google acknowledges that generative AI is a relatively new technology and therefore prone to error, and has limited its functionality to prevent misinformation about the election.
“Especially during federal and statewide elections, people rely on us to provide them with reliable, up-to-date information on topics like current candidates, the voting process, and election results. This new technology may make mistakes as it learns and as news develops,” Laurie Richardson, Google's vice president of trust and safety, wrote in a blog post.
Misinformation was a major issue in the 2020 presidential election, and with the rise of generative AI, it's poised to become an even bigger problem in the 2024 election. With changes announced on Friday, Google is trying to get ahead of the problem.
The company also announced that, as it has done in past elections, it will introduce a feature in Google Search to help people across the country find information about registering to vote.
Additionally, users will see new features on YouTube to help them find reliable information about political candidates and their parties, and as the election approaches, the video platform will start showing reminders on where and how to vote.
To help people find trustworthy election information, Google Play has introduced new badges that will appear on official government apps.