Google restricted AI chatbot Gemini from answering questions on 2024 elections

Despite efforts to block sensitive questions, controversies persist over the selection criteria for restricted queries

Google's Gemini AI chatbot is implementing restrictions on answering election-related questions in countries where voting is scheduled for this year. The decision aims to curb the dissemination of potentially misleading or biased information during critical electoral periods. Initially announced in December and subsequently enforced for European parliamentary elections in February, Google is now extending these limitations globally, as indicated by recent statements from its India team.  Gemini's responses to election-related queries have become evasive, redirecting users to traditional Google search results instead. This move comes amid growing concerns over AI-generated disinformation and its potential impact on elections worldwide.

In response to the heightened scrutiny, Google is implementing various measures, including digital watermarking and content labeling, to mitigate the spread of misinformation. However, this decision to restrict Gemini's functionality prompts questions about the broader reliability of Google's AI tools. Critics argue that if these tools are deemed unfit for providing information on democratic elections, their trustworthiness in other domains, such as health or finance, comes into question. Google's Gemini AI previously faced criticism for inaccurately generating images of people of color, leading to suspensions of certain capabilities and calls for accountability from lawmakers. The incident underscores the complex challenges faced by major AI firms, which increasingly find themselves entangled in cultural debates and subject to public scrutiny.