Google continues to limit Gemini’s ability to answer political questions, even as competitors like OpenAI, Meta, and Anthropic have adjusted their AI models to provide more nuanced responses. While many AI companies initially restricted election-related queries to avoid misinformation risks, Google has not announced any updates to its policies despite major elections passing. When asked political questions, Gemini often refuses to respond, stating it cannot help with such topics.
Testing by TechCrunch found inconsistencies, with the chatbot sometimes struggling to provide factual political information. In one case, Gemini mistakenly referred to Donald Trump as the “former president” and declined to clarify when questioned further. Google attributed the mistake to confusion over Trump’s nonconsecutive terms and stated it is working on fixing such errors.
After TechCrunch flagged the issue, Gemini began correctly identifying Trump and J.D. Vance as the current U.S. president and vice president. However, the chatbot remained inconsistent, occasionally refusing to answer similar questions.
Unlike Google, AI firms like OpenAI and Anthropic have taken steps to ensure their models engage more openly in political discussions while avoiding harmful content. OpenAI recently emphasized its commitment to intellectual freedom and avoiding censorship, while Anthropic’s latest Claude model has been designed to refuse fewer queries while maintaining accuracy.
Despite these shifts, Google seems to be taking a more cautious approach, likely to minimize potential controversies. However, critics, including some of Trump’s advisors like Elon Musk and David Sacks, argue that such restrictions amount to AI censorship.
As AI chatbots increasingly influence public discourse, the debate over balancing accuracy, freedom of information, and responsible AI governance continues to grow. For now, Google appears to be behind its competitors in adapting to the evolving expectations for AI-driven political discussions.
