Mark Zuckerberg calls Altman, Pichai chatbots ‘problematic’, says Meta is fixing it

Meta’s Llama 4 AI model is being updated to address “left-leaning” political bias, a move distinct from traditional concerns. Meta claims this bias affects all leading LLMs, including those from Google and OpenAI, due to internet training data. Llama 4 reportedly refuses fewer controversial topics and offers more balanced responses, aiming for neutrality.