New Framework Limits AI Claims to What Was Actually Tested

Original: From Protocols to Evidence: Bounded Claims for AI in Service of the Common Good

Source: arxiv.org ↗