Google keeps playing it safe with political queries on Gemini

Google’s AI chatbot Gemini seems to be playing it safe when it comes to discussing politics. While other chatbots are diving into politically sensitive topics, Gemini often dodges questions about elections and political figures, claiming it “can’t help with responses right now.” This cautious approach has raised some eyebrows, especially since last year’s major elections have already passed.

Gemini’s reluctance to provide factual political information has led to some embarrassing moments, like calling Donald J. Trump the “former president” and struggling to identify the current U.S. president and vice president. Google admits that Gemini can sometimes give outdated information or get confused by nonconsecutive terms, but they’re working on fixing these issues.

After being called out for its errors, Gemini started giving more accurate answers about the current president and vice president. However, it’s still not consistent and occasionally refuses to answer questions. Despite these hiccups, Google continues to take a cautious approach by limiting Gemini’s responses to political queries.

Some Silicon Valley figures, including Trump’s advisers, have accused companies like Google and OpenAI of AI censorship for restricting chatbot responses. In response to this criticism, OpenAI has emphasized intellectual freedom and ensuring their AI models don’t censor certain viewpoints. Anthropic has also improved their AI model to make more nuanced distinctions between harmful and benign answers.

While other AI labs are striving to provide balanced and informative responses to tough political questions, Google’s Gemini seems to be lagging behind. It remains to be seen whether Google will update its policies and catch up with the competition in the AI chatbot arena.