OCP Logo
OCP Logo

Buying AI

Tips and tools

Region

Share

Contract management for AI technology

WhoProcurement, project, and technology teams
WhatTips for contract management

While an effective procurement process and clear, simple contract can help set your internal and vendor teams up for success, the work continues in contract management.

Good contract management is critical to get the most from your AI contracts, secure good value for money, and to keep the project within scope. A lot of value can be gained or lost during the contract management phase, which is why close alignment between the procurement, project, and technology teams matters.

Contract management tips

  • Focus on communication with your internal technology and project teams. AI technology requires ongoing maintenance. This has cost, audit, and security implications. For example, with flexible services, it can be easy for an internal team member to spin up a report or monitoring service and unknowingly go over cost. Sometimes technology teams can also accidentally set off vendor security alerts by not going through proper channels with the vendor when conducting vulnerability assessments, leading to service interruptions. Open and clear internal communication and a shared understanding of boundaries with internal teams can help you keep your project within scope, maintain good vendor relationships, and maintain smoothly operating services.
  • Fit your vendor management approach to the size and scope of the contract. How often you check in with a vendor will vary. Buying development services is different from buying access to a service, and a large, high-profile project is different from a simpler and lower-risk project. No matter the scope, your check-in meetings should help gauge if you are continuing to get good value from the contract.

Contract management: Key questions

Procurement objectives

  • Manage contract scope
  • Ensure prompt payment

Organizational objectives

  • Implement and refine the AI solution
  • Conduct change management and train staff to support solution adoption and scaling
  • Prepare for decommissioning
TeamQuestions
Procurement

Operational considerations

  • Is the vendor meeting our milestones?
  • Are we paying the vendor in a timely manner?
  • Do we need to issue a change order?

Innovation and scalability

  • How will we support scaling what works, while allowing flexibility for departments to innovate?
  • What structures will support knowledge sharing and solution reuse across teams or regions?
  • How will we stay informed about AI capabilities and vendor offerings?
Agency or department buyer

Operational considerations

  • What training and support will we offer to end users, especially in high-risk or public-facing roles?
  • How will we test and evaluate deliverables during the development or implementation phase? Who will be responsible for these activities?
  • How satisfied are users?
  • How will we document and share AI learnings and best practices across the organization?
  • Is the payment model working out as expected?

Innovation and scalability

  • How will we enable safe experimentation (e.g., pilots, sandboxes) while maintaining appropriate controls?
  • How will we engage with the wider public sector, academic, or civic tech community to share best practices, create spaces for reflection and learning, or co-develop tools?
IT/Data analytics

Innovation and scalability

  • What frameworks will we use to evaluate, iterate on, and scale successful AI pilots?
  • How can we support interoperability between different AI platforms or ecosystems over time?
  • Is there room for improvement in the solution?
  • What could have been done differently?