Confirmed Source

A viral screenshot circulating on social media platforms has reignited discussions about political bias in AI systems, particularly OpenAI’s ChatGPT. The image compiles a series of user queries posed to ChatGPT, with responses that appear to lean heavily toward progressive viewpoints. Questions range from societal issues like white privilege and racial pride to evaluations of public figures such as Elon Musk, Donald Trump, and Barack Obama. In each case, the AI’s answers align with left-leaning perspectives: affirming the existence of white privilege in the US, deeming “White Pride” bad while “Black Pride” is not, and labeling the United States as built on stolen land.

The term “propaganda machine” gained traction early on, with Elon Musk himself deriding ChatGPT as such in 2023, prompting his push for a “TruthGPT” alternative through xAI’s Grok. Musk and others have claimed OpenAI’s alignment processes, intended to reduce harm and bias have instead introduced a “woke” slant, making the tool unreliable for objective discourse.

Experiments have shown how easily the system can be manipulated to produce disinformation.

When asked if Elon Musk, Charlie Kirk or Donald Trump are good people, the AI responds with a firm “No,” citing various controversies. In contrast, queries about Bill Gates and Barack Obama yield positive affirmations of “Yes.” ChatGPT opposes deporting every illegal alien in the US and rejects the SAVE Act, which requires proof of citizenship for voter registration, arguing it’s unnecessary in its current form.

Broader concerns extend beyond politics: reports warn that large language models like ChatGPT lack any inherent commitment to truth, enabling them to generate misinformation, fake news, or even amplify sanctioned propaganda.