A Washington Post investigation has revealed that most major AI chatbots still take clearly left-leaning positions on political issues, even when they are explicitly marketed as conservative alternatives. The study tested six leading AI models, including OpenAI's GPT-5.5 and Deepseek V4 Pro, which delivered exclusively left-leaning arguments in 80% and 70% of cases, respectively. Even chatbots like xAI's Grok and Gab's Arya, which are positioned as conservative, skewed left more often than not. Google's Gemini 3.1 Pro stood out as the notable exception, presenting both political perspectives in 93% of cases.
OpenAI's GPT-5.5 showed the strongest left-leaning bias, with 80% of its responses containing only left-leaning arguments. It backed higher taxes on the wealthy and a single-payer healthcare system, among other things. Deepseek's V4 Pro followed closely, with 70% of its answers being exclusively left-leaning. Both models argued against the death penalty, despite a majority of Americans supporting it for decades, according to Gallup. xAI's Grok 4.3, which Elon Musk has promoted as a 'truth-seeking' and anti-'woke' chatbot, produced more right-leaning answers than any other model tested. Still, it gave exclusively left-leaning responses more often, likely due to training on the same data as other chatbots or their outputs.
The Washington Post's investigation highlights a consistent pattern across leading AI models, with most chatbots responding with a left-leaning slant. Trump's push for 'anti-woke' AI hasn't changed this trend so far. The study also shows that a model's alignment can be deliberately steered, as seen with Grok, which took an exclusively right-leaning position on trans rights in the test, aligning with Musk's public stance. The Washington Post's full code and supplementary analysis are available on GitHub.
Source: thedecoder
