The World Economic Forum (WEF) has made available a new set of guidelines describing how to best procure AI-related systems and services, providing a handy toolkit aptly entitled AI Procurement In A Box .
Aimed primarily at the public sector to boost best practices for procuring AI by governmental entities, the guide equally offers insights and approaches that directly apply to private industry. By bolstering how AI is procured there is a resolute hope that those crafting AI will be prodded toward ensuring that their AI systems are trustworthy and conform with appropriate AI Ethics principles.
The AI Procurement In A Box methodology is structured into four overarching modules and the second module contains a detailed workbook.
Here’s a summary of the key components:
- Module 01: Guidelines for government AI procurement – provides an overview of procuring AI-powered solutions
- Module 02: Workbook for policy and procurement officials – contains workbook details and encompasses tools for both thinking and rethinking about how to approach AI procurements
a. Risk assessment
b. User Manual
c. AI specification and evaluation tool
d. Workshop slide pack: How to kick-off the implementation
e. Case studies - Module 03: Challenges and opportunities during implement – showcases insights gleaned from workshops that occurred when initially putting the guidelines into practice
- Module 04: Pilot case studies from the United Kingdom – indicates case studies depicting the use of the AI procurement guidelines by U.K. entities
As noted by Kay Firth-Butterfield, Head of AI and Machine Learning and a member of the Executive Committee for the WEF, these vital guidelines will help spur the adoption of agile ways to have AI benefit humanity and the planet, while simultaneously seeking to mitigate adverse negativities from AI. Via the use of the provided procurement methods, governments have the potential to judiciously kick-start a budding AI economy within their respective countries.
Reporting by Dr Lance Eliot our US Associate Editor for Computers & Law
