Are AI chatbots like ChatGPT politically biased? We tested them.

Political bias in AI chatbots is a contentious issue, prompting President Trump’s executive order on neutral, nonpartisan AI tools.
The Washington Post tested AI models behind major chatbots (e.g., OpenAI’s ChatGPT, Google’s Gemini) using political questions and found that most chatbots have left-leaning political biases.
The ChatGPT model predominantly offered left-leaning responses, while Google’s Gemini mostly adopted a balanced, both-sides approach.
Even AI systems marketed with conservative values, like Elon Musk’s Grok, often leaned left in their responses, according to The Washington Post’s analysis.
Experts argue that these AI tools are becoming influential in shaping public perception, and understanding their inherent biases is crucial for ensuring neutrality in their responses.
The bias in chatbots’ political responses has been consistent across multiple academic studies, and these tools continue to contribute a non-neutral representation of political debates.
The decisions made by tech companies in developing these AI tools, including the data included and the programming guiding their responses, have potential political implications.

You may have missed