Google said on Tuesday that its AI chatbot Gemini won’t give answers about upcoming global elections to reduce the risk of spreading wrong information.
This comes as people worry about how AI technology might be misused, especially during elections when fake news can spread easily.
Gemini, made by Google’s parent company Alphabet, will no longer answer questions about elections, including big events like the upcoming U.S. presidential election between Joe Biden and Donald Trump. Instead, people will be told to use Google Search.
These restrictions were first announced for the U.S. in December, but now they’re for the whole world as different countries get ready for elections in 2024. Places like South Africa and India, which have the biggest democracy, are having important elections.
In India, the government wants tech companies to get permission before releasing AI tools, especially if they might give out wrong or unverified information.
Google CEO talks about biases in Gemini’s responses (Via Cameron Stiller/Shutterstock)
Google decided to limit Gemini’s answers after people found mistakes in the chatbot’s picture-making feature and stopped it for a while.
Google’s CEO, Sundar Pichai, admitted there were problems with Gemini’s answers, saying they were biased and not okay. The company is trying to fix these issues and make its AI products more trustworthy.
This move is part of a bigger effort in the tech industry to fight against spreading wrong information, especially before important political events like elections.
Meta Platforms, the company that owns Facebook, said it will set up a special team to deal with wrong information and misuse of AI tools as the European Parliament elections get closer in June.
