Google Restricts Gemini From Answering Election-Related Queries Worldwide

Google has begun to restrict the types of election-related queries Gemini will return in its responses.

The company announced the changes on Tuesday. Restriction in India is one of the latest changes.

Google pulled its artificial intelligence (AI) multimodal tool last month following a string of controversies, including historical inaccuracies of race.

Technology platforms continue to prepare for a year of controversial elections in the United States, as well as elections worldwide that will affect billions of people in more than 40 countries.

The rise of AI-generated content has led to serious election-related misinformation concerns, with the number of generated deepfakes increasing 900% year-over-year, according to data from machine-learning firm Clarity.

According to data from Civic Science, Republicans are more likely to be concerned about the influence of deepfakes on the 2024 election. Severe concern has fallen by eight percentage points among Democrats, although 50% still hold strong concern.



Critics say detection and watermarking technologies used to identify deepfakes have not advanced enough to keep up. Some are not confident that platforms have the technology to protect images and videos.

Google wrote in a blog post published today that the company is collaborating with the Election Commission of India (ECI) to enable people to easily discover critical voting information on Google Search, including information on how to register and how to vote. The information will appear in English as well as Hindi.

Google is not the only company attempting to self-regulate AI systems. In mid-February, two dozen technology companies working on AI said they signed a pledge to try to prevent their software from interfering in elections. Many are located in the United States.

These companies include Adobe, Amazon, Google, Meta, Microsoft and X. The list also includes startups that allows people to make fake voices generated by generative AI (GAI) product that people could use to abuse election information.

The voluntary accord does not ban AI content in elections, but it does list steps that the companies say they will take, including the development of tools, to identify generative artificial intelligence (GAI)-generated images from authentic content.

Next story loading loading..