With artificial intelligence technology fuelling digital procurement transformation at an unprecedented pace in recent years, the World Economic Forum’s Centre for the Fourth Industrial Revolution has just published a set of 10 AI-purchasing guidelines specifically aimed at governments, which it says must adopt a “special” approach to AI procurement that differs from practices common in the private sector.
The Forum’s global network on technology governance prepared the white paper, which warns of the potential “negative consequences” of AI.
The guidelines state: “Governments do not have the latitude of using the inscrutable ‘black box’ algorithms that increasingly characterize AI deployed by industry.”
Whereas private enterprises using AI algorithms can claim commercial sensitivity to protect against third-party scrutiny, public agencies have a duty to provide “accountability, transparency and explainability” in order to maintain public confidence and avert the danger of creating inadvertent new risks or causing harm, the guidelines emphasise.
The guidelines favour governments that are adopting new approaches to AI-powered procurement that draw upon multi-disciplinary teams on both sides of the procurement equation, with the aim of building a level playing field for all suppliers.
An iterative approach features prominently: “Use procurement processes that focus not on prescribing a specific solution but rather on outlining problems and opportunities, and allow room for iteration.”
Public sector procurement teams should, the document proposes, collaborate and learn from each other’s experience in AI procurement, all the while seeking to align their buying processes with relevant national strategies focused on emerging technologies.
Bidders should be made aware from the outset that public benefit is the purpose of these procurement efforts, and the guidelines urge mandatory risk and impact assessments at the start of the procurement process and following key decisions.
Transparency is clearly a priority, with the white paper insisting that the focus throughout the AI-procurement process should be on implementing effective mechanisms for “algorithmic accountability and transparency norms”.
The British Government is piloting all ten of the white paper’s guidelines.