Today, I submitted comments on proposed GSAR 552.239-7001, the “Basic Safeguarding of Artificial Intelligence Systems" clause that GSA intends to incorporate into all MAS contracts. This clause, as drafted, suffers from several serious deficiencies that will undermine the very objectives it seeks to achieve, as well as President Trump’s broader AI policy objectives. Some of these are technical and can be remedied with careful revision. Others reflect deeper conceptual confusions that, if left unaddressed, will chill participation by the most capable AI providers, concentrate the government’s AI supply chain in firms selected for political compliance rather than technical excellence, and establish precedents that future administrations—of any political persuasion—will exploit in ways the current drafters may not intend.
