Google is limiting its AI chatbot from answering questions related to elections to prevent the spread of misinformation.
When asked about elections or anything related to voting, Google’s AI chatbot Gemini will now prompt users to search it on Google. It’s the latest attempt to prevent AI tools from manipulating voters.
Google is restricting its AI chatbot to answer questions related to elections in countries where elections are taking place this year to avoid spreading misinformation.
If you ask Gemini a question related to elections, politicians, or political parties, it now responds with: “I’m still learning how to answer this question. In the meantime, try searching it on Google.”
A Google spokesperson told Sky News that the restrictions were introduced “with regard to the many elections taking place worldwide in 2024, and out of abundant caution.”
In February, Google halted Gemini from generating images after it created a series of inaccurate historical depictions of individuals.
The model was supposed to reflect a diverse range of people but had become “much more cautious than we had intended,” said Prabhakar Raghavan, Senior Vice President of Google.
This year, elections are taking place in more than 50 countries. With the increasing capabilities of artificial intelligence, concerns are growing that it could be used to manipulate voters.
Just two days before the elections in Slovakia last September, a fake audio recording was released on Facebook.
It sounded like one of the candidates and a journalist discussing how to manipulate the elections. The recording was quickly recognized as fake, generated by AI, but that didn’t stop it from spreading.
The candidate narrowly lost the election.
Now, technology companies and governments are increasingly cautious ahead of elections.
Meta, which owns Facebook, Instagram, and WhatsApp, is forming a team to combat misinformation and the misuse of artificial intelligence ahead of the European elections in June.
In India, companies have been asked to seek government approval before publicly releasing AI tools that are “unreliable” or in the experimental stage and to label them as potentially incorrect responses.
This step came after journalist Arnab Ray asked Google’s Gemini if the Prime Minister of India was a fascist.
He says Gemini responded that Mr. Modi “is accused of implementing policies that some experts have termed fascist.”
A government minister then accused Google of violating India’s information technology laws.
Junior Minister for Information Technology Rajeev Chandrasekhar wrote on X that the unreliability of AI platforms cannot serve as an excuse for them to be exempt from Indian laws.
India is expected to go to the polls between April and May.
Discussion about this post