Anthropic Evaluates Political Neutrality in AI Model Claude

Anthropic unveils a new evaluation method assessing political neutrality in AI models, with Claude Sonnet 4.5 outperforming competitors like GPT-5 and Llama 4.