In a bid to prevent potential missteps in the deployment of its technology, Google has announced restrictions on its AI chatbot Gemini from responding to queries about global elections scheduled for this year, the Alphabet-owned firm revealed on Tuesday. This decision comes in the wake of growing concerns surrounding misinformation and fake news, particularly amidst advancements in generative AI technologies like image and video generation, prompting heightened scrutiny and regulatory actions by governments worldwide.
Gemini’s response to inquiries regarding elections, including the forthcoming U.S. presidential match-up between Joe Biden and Donald Trump, now prompts users with a message stating, “I’m still learning how to answer this question. In the meantime, try Google Search.” Google had initially announced such restrictions within the U.S. in December, with plans for their implementation prior to the election.
“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,” disclosed a company spokesperson on Tuesday. Besides the United States, national elections are slated to occur in several significant countries, including South Africa and India, the world’s largest democracy.
India, in particular, has taken proactive measures by requiring tech firms to obtain government approval before releasing AI tools that are deemed “unreliable” or are still in the trial phase, mandating clear labeling to caution against potential inaccuracies. Google’s AI products have come under scrutiny following recent inaccuracies in historical depictions generated by Gemini, leading to the temporary suspension of the chatbot’s image-generation feature last month.
CEO Sundar Pichai acknowledged these issues, labeling Gemini’s responses as “biased” and “completely unacceptable,” while assuring that corrective measures were underway. Concurrently, Facebook-parent Meta Platforms announced plans to establish a dedicated team to combat disinformation and the misuse of generative AI in the lead-up to the European Parliament elections slated for June, underscoring the industry-wide efforts to address the evolving challenges posed by AI-driven misinformation campaigns.