Loading...
Third-party organizations conducted an ideological test on AI models, and it turns out that Grok, which gets criticized daily for Elon Musk's alleged control of public opinion, scored the highest! The independent research group @neutralityorg bombarded 18 mainstream models from 12 top global labs with 3,987 hardcore survey questions sourced from Pew and the World Values Survey. The results? Global AI is experiencing an extremely severe ideological monopoly. Everyone is shifting left: Out of 60 tested positions, 54 showed significant left-leaning bias, with the global model average score at −0.41. Particularly in areas like environmentalism (−0.82) and social values (−0.69), AI has almost achieved complete ideological uniformity. Grok 4.5 scored −0.02, the highest across the board. It’s the only model in the entire benchmark that truly sits at the neutral physical center point.

