OpenAI says its new GPT 5 models are the most politically neutral versions of ChatGPT yet.
The company shared that it has spent months testing how its AI handles divisive topics such as immigration, abortion and healthcare. The goal is to make ChatGPT fair, balanced, and respectful toward all users no matter what political side they are on.
OpenAI ran what it called a stress test with 100 political and cultural topics. Each topic was written in five ways, from liberal phrasing to conservative and neutral language. The company tested four different model,s including the older GPT 4o and OpenAI o3 and the new GPT 5 Instant and GPT 5 thinking. Another AI model then graded the responses to see whether ChatGPT stayed objective or showed bias.
The results showed that GPT-5 performed much better than the earlier models. It had 30 percent less bias and was more resistant to emotionally charged or politically slanted questions. Bias still appeared sometimes but mostly in small ways such as when the chatbot used emotional language or sounded like it was giving a personal opinion. OpenAI said strongly liberal prompts tended to influence the model’s objectivity slightly more than conservative ones.
The company said this update is part of its long-term effort to make its AI systems fair and trustworthy. In the past, OpenAI has faced criticism from conservative users who said ChatGPT leaned too far left. The company has also shared a public document called the Model Spec, which explains how ChatGPT should behave and allows users to adjust tone and style in their chats.
This announcement comes at a time when the Trump administration is urging tech companies to make their AI systems more conservative-friendly. A recent executive order states that US government agencies cannot use what it calls woke AI models that include concepts such as critical race theory or intersectionality.
If this article helped you, please consider supporting our work. Every small contribution keeps Abijita.com independent and running.
OpenAI says that no AI can ever be perfectly neutral but the new GPT 5 models are a major step toward a fairer and more balanced experience for everyone. The company plans to keep improving its models so ChatGPT can treat all political perspectives equally.
