Leading AI models (even Grok) are all a bunch of leftist punks

Researchers examining several prominent artificial intelligence models have identified a consistent tendency toward progressive or liberal viewpoints in generated responses. This study highlights ongoing challenges in training systems to provide neutral outputs across diverse political topics, raising questions about how inherent data biases influence the neutrality of automated information tools.