X, formerly known as Twitter, has introduced an AI-powered enhancement to its Community Notes feature, allowing large language models (LLMs) to draft fact-checking notes on posts. These AI-generated notes, termed 'AI Note Writers,' are designed to accelerate the fact-checking process and expand the reach of the platform’s contextual information on viral content. Despite the automation, human contributors retain final approval authority, as AI-generated notes will only be published if rated as helpful by diverse human reviewers. This hybrid approach aims to address the slow approval rates that previously allowed misinformation to persist, with humans historically approving only about 3% of 52,000 note proposals. Additionally, X has launched an AI Note Writer API, enabling developers to create and submit their own AI fact-checking agents for review. The move has sparked discussions about the reliability of AI in fact-checking, given concerns over potential hallucinations or inaccuracies by language models. X’s vice president of product has highlighted the intention to maintain human oversight to prevent AI from promoting misleading fact-checks, ensuring the credibility of Community Notes remains intact.