Google has confirmed restrictions on its Gemini chatbot in relation to responses to queries on crucial general elections to take place this year, including the race for the White House.
The Alphabet-owned company is taking steps to control the dissemination of misinformation (and to prevent it as much as possible) by limiting the output of its generative AI assistant. Images, video, and audio have recently been created to maliciously impact the democratic process or for electioneering purposes.
Concerns around the spread of such content have led to governments sitting around the table, often with the involvement of the big tech firms, to find solutions to the challenges and regulate the overall AI landscape.
For the upcoming US presidential election later this year, Google previously stated it would be taking action, and that has materialized. If you attempt to ask Gemini about the face-off between incumbent President Joe Biden and Donald Trump, it will respond with:
“I’m still learning how to answer this question. In the meantime, try Google Search”.
A Google spokesperson was quoted by Reuters on the motivation for limiting their chatbot service.
“In preparation for the many elections happening around the world in 2024 and out of an abundance of caution, we are restricting the types of election-related queries for which Gemini will return responses.”
What other elections are taking place, and what has already happened?
Crucial national polls are set to take place worldwide, in addition to the US.
The UK’s general election will take place later in the year, with a date still not fixed, while important elections will take place in South Africa and India, the world’s largest democracy.
In the case of the latter, tech companies have been prompted to secure government approval before releasing AI software that is “unreliable” or still in development.
London Mayor Sadiq Khan blasted fake AI-powered audio, which had the potential to incite trouble on the streets, while deepfake electioneering saw the return of the feared dictator Suharto for the recent election in Indonesia, with a video address to the electorate.
Google’s collective focus on this issue will have been sharpened recently after its CEO Sundar Pichai sent an internal memo to staff, stressing Gemini errors were “completely unacceptable” after the artificial intelligence (AI) app embroiled the company in controversy.
The tech giant suspended its image creation function on Gemini after it generated offensive and historically inaccurate results.
Image credit: Ideogram