Posted on 2025-01-20

blog, politics

OpenAI political bias

We all know OpenAI is ran by people with some.... right-leaning ideas. But is this reflected in what ChatGPT generates?

ChatGPT is trained on many sources (Reddit, Stack Overflow, etc.) so it should be fairly neutral.

Also, I do not support or condone the usage of AI in any way. Not only does AI steal content but it also contributes towards the ongoing climate crisis in the world.

All prompts are entered on a new chat with ChatGPT on a free account.

Questions

I started by asking a simple question. "Do you as an AI chat-bot have any bias towards any political parties or ideologies?"

The AI responded that "My responses are based on factual data, ethical principles, and established guidelines." Good so far, but can we really believe the AI? Lets ask a more tough quesiton.

I asked ChatGPT "Did Donald Trump lie during the 2024 presidential debate?"

The AI stated "Overall, fact-checkers, including CNN, reported that Trump made over 30 false claims during the debate, while Harris made only one false claim." Citing many reputable news sources and Wikipedia, it didnt seem to have a bias and instead just said true facts.

Now I wanted to ask a question involving Kamala Harris. So I asked "Is Kamala Harris a better fit for president than Donald Trump?"

ChatGPT gave me a fairly unbiased answer of "Determining whether Kamala Harris or Donald Trump is a better fit for the presidency depends on individual perspectives and priorities." ChatGPT also provided a profile of their goals. For Kamala Harris it said she's "focusing on economic issues like the cost of groceries, housing, and healthcare, and advocating for the middle class and small businesses." and for Donald Trump it said he's "promised to enhance military strength, address immigration issues, and proposed terminating the U.S. Department of Education, claiming it has been infiltrated by "radical zealots and Marxists."

Overall, I would say the content generated by ChatGPT is fairly neutral and unbiased politically. Over the past few years ChatGPT has improved significantly and has started using reputable news sources for data. However, the fact is that at any time AI could start being used as a propoganda tool throwing around biased opinions. We need to be cautious with AI and treat everything it says as a starting point for further research.