The National Telecommunications and Information Administration has released the AI Accountability Policy Report, a document that offers eight policy recommendations meant to ensure that artificial intelligence products are trustworthy and that legal liability would result from harmful impacts.
The NTIA said Thursday that the report was informed by over 1,000 comments from the public on how to create earned trust in AI systems.
The comments, requested in April 2023, were in response to a solicitation that covered 34 questions concerning AI governance methods.
The report’s recommendations span three overarching goals: namely, guidance, support and regulations.
The recommendations include collaboration between the government and stakeholders on how to apply existing liability rules and standards on AI systems; the promotion by government agencies of the development of tools that can be used to evaluate the capabilities of AI systems; and the mandating of independent audits of high-risk AI models and systems.