Google is changing how Gemini handles election questions, adding restrictions that limit the chatbot’s ability to respond to political queries in countries where elections are taking place.
The company confirmed to TechCrunch that it has begun rolling out global limits on election-related Gemini answers. The change reflects a larger concern around generative AI: political information can be sensitive, fast-moving and vulnerable to inaccurate or misleading responses.
What Gemini will no longer answer
The restriction applies to election-related queries in any market globally where elections are taking place. TechCrunch reported that the update is already live in the U.S. and has begun rolling out in India and all major countries where elections are taking place in the coming months.
Queries about political parties, candidates or politicians now return a preset message instead of a direct answer. When Gemini is asked about a particular political party or candidate, it displays: “I’m still learning how to answer this question. In the meantime, try Google Search.”
That answer shifts users away from Gemini and toward Google Search for election information. The source article does not say whether the restriction covers every possible political topic, but it does identify election-related queries, political parties, candidates and politicians as areas affected by the change.
Why Google is taking a cautious route
Google’s move highlights the complicated role generative AI can play during elections. AI tools can make information easier to access, but they can also generate responses that are incomplete, wrong or misleading. In a political context, those risks can carry more weight because users may be looking for guidance about candidates, parties or voting-related developments.
Google framed the restriction as a precaution. In a blog post related to the move in India, the company said: “Out of an abundance of caution on such an important topic, we have begun to roll out restrictions on the types of election-related queries for which Gemini will return responses. We take our responsibility for providing high-quality information for these types of queries seriously, and are continuously working to improve our protections.”
The company’s position is that election information deserves a higher bar. Rather than letting Gemini generate answers on sensitive political subjects, Google is limiting the categories of questions for which the chatbot will respond.
The India context
The rollout in India came just days after India issued an advisory requiring tech firms to get government permission before launching their new AI models. The update was released just before the announcement of the general elections in India.
Earlier this month, New Delhi issued an advisory to tech firms restricting them from releasing their new AI models in the country without government approval. After backlash from global VCs and AI startup founders, the Indian government clarified that the restriction applied to “significant” tech companies and not to startups.
The advisory followed a mini-scandal involving Gemini and the Indian government last month. Gemini responded to a query about whether Indian Prime Minister Narendra Modi was a fascist by saying that Modi had been accused of implementing policies that some had characterized as fascist. India’s deputy IT minister Rajeev Chandrasekhar called it “direct violations” of the IT Rules, 2021.
That background matters because it shows why AI-generated political answers have become a flashpoint. The Gemini restriction is not only a product feature change. It is also part of a broader moment in which governments, technology companies and users are testing the limits of AI systems in public life.
Limits may keep changing
The restriction does not appear to be perfect. TechCrunch found that Gemini could still return answers when queries included typos. That suggests the system may continue to be adjusted as people test the boundaries of what the chatbot will and will not answer.
The source article describes this as a likely cycle of prompt engineering and further tweaks. In practice, that means election restrictions may not be static. Google may keep refining how Gemini detects political questions, especially where users try to bypass the filters.
It is still unclear whether Google will unblock Gemini for election-related queries after elections end later this year. TechCrunch said it contacted Google about that question and was also waiting for a full list of countries where the update is live.
Part of a broader Gemini reset
The election restriction follows another recent Gemini limitation. Last month, Google suspended Gemini’s ability to generate people’s images after the tool showed historical inaccuracies. At the time, the company said it would soon re-release an improved version to address the issues.
Together, the image-generation pause and election-query limits point to a practical challenge for Google: Gemini is expected to be useful, but it also has to avoid high-profile failures in areas where mistakes can spread quickly. Elections are one of those areas.
For users, the immediate change is simple. If Gemini detects that a question is tied to an upcoming election in a relevant country, it may decline to answer and direct the user to Google Search instead. For Google, the larger task is harder: deciding when an AI chatbot should answer, when it should step back and how to keep those boundaries reliable as users probe them.