Cary Coglianese (University of Pennsylvania Law School) and Erik Lampmann (University of Pennsylvania Law School) have posted “Contracting for Algorithmic Accountability” (Administrative Law Review Accord, vol. 6, p. 175, 2021 on SSRN. Here is the abstract:
As local, state, and federal governments increase their reliance on artificial intelligence (AI) decision-making tools designed and operated by private contractors, so too do public concerns increase over the accountability and transparency of such AI tools. But current calls to respond to these concerns by banning governments from using AI will only deny society the benefits that prudent use of such technology can provide. In this Article, we argue that government agencies should pursue a more nuanced and effective approach to governing the governmental use of AI by structuring their procurement contracts for AI tools and services in ways that promote responsible use of algorithms. By contracting for algorithmic accountability, government agencies can act immediately, without any need for new legislation, to reassure the public that governmental use of machine-learning algorithms will be deployed responsibly. Furthermore, unlike with the adoption of legislation, a contracting approach to AI governance can be tailored to meet the needs of specific agencies and particular uses. Contracting can also provide a means for government to foster improved deployment of AI in the private sector, as vendors that serve government agencies may shift their practices more generally to foster responsible AI practices with their private sector clients. As a result, we argue that government procurement officers and agency officials should consider several key governance issues in their contract negotiations with AI vendors. Perhaps the most fundamental issue relates to vendors’ claims to trade secret protection—an issue that we show can be readily addressed during the procurement process. Government contracts can be designed to balance legitimate protection of proprietary information with the vital public need for transparency about the design and operation of algorithmic systems used by government agencies. We further urge consideration in government contracting of other key governance issues, including data privacy and security, the use of algorithmic impact statements or audits, and the role for public participation in the development of AI systems. In an era of increasing governmental reliance on artificial intelligence, public contracting can serve as an important and tractable governance strategy to promote the responsible use of algorithmic tools.