Who to involve

Procuring complex technology like AI may require coordination and input across a range of stakeholders. Off-the-shelf technology can require less engagement, depending upon the use case. But for customized or built AI, or higher-risk use cases for off-the-shelf technology, it’s important for stakeholders to be involved as soon as you can, and often throughout the deployment process. Bringing together people with diverse backgrounds, perspectives, skills, and experiences can help your organization reflect on the overall impact and how to make the most from AI, while also considering requirements for security, privacy, explainability, audibility, and legal compliance.

It’s especially important to bring in your technology experts into your conversations as early as possible to help you understand if AI is a feasible and valuable tool to address your challenge, as well as navigate the technical components of the procurement and implementation processes. If you lack in-house expertise, consider what experts from your community and nonprofit partners or other experienced buyers you can consult, and think carefully about what solutions you can feasibly pursue given your capacity constraints.

Precisely whom you will involve will be different depending on your specific context. Your local laws, procurement system (centralized versus decentralized), and the specific purchase type will all impact your own final list of whom to engage.

Key roles include:

• Procurement officer, who leads the procurement process. They should help articulate the scope and strategy of the procurement, and often have final signoff on the vendor selection process, and then provide oversight during contract implementation.
• Project team leader (buyer), or the agency/department representative that is seeking to make a purchase to meet a need. They usually facilitate the overall discovery process, ensuring smooth collaboration and progress. In a best-case scenario, this person is a product owner, who also has technology expertise.
• IT/Data science expert, who can provide knowledgeable guidance around responsible AI use and its capacities and limitations, as well as infrastructure and security requirements, and more. They are usually in an advisory role, unless they are also in the position of being the buyer.
• Legal officer, who can help navigate considerations around data, liability, intellectual property, and AI use cases. They often provide sign-off on both the RFP and final contract.

Two other critical roles are the advocate or representative of the end users, and the public. Although they are different types of stakeholders, it’s still important to consult them and gather their feedback along the way. Public participation is especially important if the AI will have a direct impact on service delivery for residents.