Photo by cottonbro studio on Pexels

A newly proposed bill aims to increase transparency in the artificial intelligence industry by requiring developers to disclose key information about their AI models. Under the proposed legislation, AI developers who choose to participate would need to make public their training data sources, evaluation methodologies, and identified limitations or failure modes. In return for these disclosures, developers would receive a ‘safe harbor’ provision, offering protection from potential lawsuits related to their AI’s performance. The bill also mandates that documentation remain up-to-date, with updates required within 30 days of any model version changes, ensuring ongoing accountability and transparency.