AI Procurement Participation Principles

Principles for Public Participation in the Procurement of AI

Version 1

Over 2024/25, ParticipationAI undertook an inquiry driven by the question: “How do we ensure meaningful public participation in decisions about data & AI?”. This led us to procurement, and a recognition that:

  • AI in public administration is more than a tool - it becomes infrastructure.
  • Current procurement models are not fully aligned with the distinctive characteristics and requirements of AI systems.
  • Stakeholder engagement — especially from civil society and affected communities — remains underdeveloped and inconsistently applied.

These principles, developed from that inquiry, and through a series of open consultations run in partnership with Connected by Data, are put forward to address some of these gaps.

We invite you to explore how you might apply these principles in your work and to sign-up for updates about using the principles and opportunities to contribute to their ongoing development.

Get in touch: contact@p4ai.net 

Preamble

Principle 1: Purpose

Principle 2: People

Principle 3: Participation at all stages

Principle 4: Planet

Principle 5: Informed engagement

Principle 6: Inclusive process

Principle 7: Transparency and oversight

Principle 8: Institutional structures

Principle 9: Market development

Principle 10: Evaluation & learning

Principles

Preamble

Introducing Artificial Intelligence (AI) into public services impacts on how institutions function, how rights are exercised and how services are experienced. AI is not just a tool, it acts as an infrastructure for future action, with long-term societal and environmental consequences. 

Specifying, appraising investment plans, buying and deploying AI involves political choices. These must prioritise  long-term ethical, social, and democratic values over short-term cost, efficiency and innovation concerns.

We are committed to shaping our use of AI, and the wider market for AI, to deliver public value by taking democratic, people-centered, anticipatory and participatory approaches to AI procurement and contracting.

We will take concrete steps to apply the following principles in our work.

Principle 1: Purpose

Begin by defining a clear public mission in partnership with the public — Every planned procurement that asks for, or considers, AI-based solutions should have defined goals grounded in evidence, societal needs, and stakeholder engagement, which should all be established before a procurement process takes place.

Consideration of AI should be based on broader, mission-oriented agendas such as reducing inequality, improving intergenerational fairness and wellbeing, strengthening democracy, and addressing the climate crisis. AI procurement and contracting choices should not be driven by vendor offerings, technology hype, ignorance of what already exists for reuse, or a predetermined decision that AI is the solution.

Standards

Supporting resources

  • We have clearly stated the problem we are trying to solve with this procurement.

  • If we are responding to an approach from vendors or advocates of AI systems, we have clearly described the problem that an AI solution addresses, validated that it is a priority for us to address, and engaged with stakeholders to consider alternative solutions.

Suggest resources or case studies in support of this principle here.

Principle 2: People

Include the right people from the start and throughout the procurement and public contracting journey –Affected communities, public servants, frontline workers, civil society, academics, technologists, and underrepresented voices may all have contributions to make from the very beginning of the procurement cycle, and should have opportunities to meaningfully engage throughout.

Revisit the question ‘Who will this procurement impact on, and how?’ at each stage of the procurement process and reach out to additional participants where necessary.

Standards

Supporting resources

  • We have identified the groups (citizens, residents, workers etc.) who would be particularly affected by this procurement and its resulting contract.

  • We have identified ways to gather evidence and input from individuals and representatives from affected groups - and considered the support they may need to engage in the procurement process.

  • We consider the interests of those who cannot easily speak for themselves, including future generations and the natural world.

Suggest resources or case studies in support of this principle here.

Principle 3: Participation at all stages

Practice participation across the entire procurement lifecycle - From needs assessment and investment appraisal to design, tender evaluation, deployment, monitoring and review - there are practical ways for stakeholders to engage, meaningfully inform and share in decision making: leading to improved market readiness, optimised specifications, better procurement and contracting outcomes.

To realise the best outcomes, public participation should be considered, meaningful and adequately resourced.

        

Standards

Supporting resources

  • We have identified appropriate methods for engagement at each stage of procurement.

  • We have identified resources to support public participation.

  • We can demonstrate how public input has made a difference to the procurement process.

Suggest resources or case studies in support of this principle here.

Principle 4: Planet

Align procurement with policy and public service environmental and climate goals - Ensure procurement is done within planetary boundaries and principles of sufficiency and equity, complementing rather than disrupting policies aimed at ensuring equitable access to resources like water, land and energy (ie. water access, housing, net zero policies).

Standards

Supporting resources

  • We include public interest goals related to our environment, resources and climate in the procurement of AI, ensuring both complement rather than contradict each other.

Suggest resources or case studies in support of this principle here.

Principle 5: Informed engagement

Build capacity for meaningful engagement around AI – For AI to be used in ways that increase government and citizen agency, procurement processes must recognise the need to invest in capabilities, not just tools.

All stakeholders—especially frontline workers and communities—need access to the knowledge and resources that will enable them to engage critically with AI systems at each stage of the procurement and deployment process.

Standards

Supporting resources

  • We have identified the technical, social and organisational knowledge relevant to make informed decisions about our proposed procurement.

  • We provide stakeholders with context-appropriate access to vendor-neutral learning opportunities about AI and its implications.

Suggest resources or case studies in support of this principle here.

Principle 6: Inclusive process

Embed equity and diversity by design - Procurement processes must include safeguards against bias and harm. Systems procured, and the processes that select and continuously improve them throughout contracted service delivery, must reflect and serve diverse populations.

Standards

Supporting resources

  • We pay particular attention to hearing voices from minoritized, atypical and excluded communities.

  • We actively discuss potential biases or harms from proposed AI solutions with stakeholders, and identify appropriate actions to manage these.

Suggest resources or case studies in support of this principle here.

Principle 7: Transparency and oversight

Ensure transparency and oversight - Audits, external reviews, and public scrutiny are all important to make sure the process of procurement, and the resulting products or services governed by public contracts, meet the original goals and achieve the anticipated outcomes and benefits set out.

As part of procuring and contracting AI, create or update comprehensive public registers of AI systems giving  clear, accessible information to the public and other public sector organisations about where AI is (or is intended to be) in use.

Standards

Supporting resources

  • We publish clear documentation on AI-related procurement processes, and on AI systems that are in  operation.

  • We provide clear and accessible information on how stakeholders have been involved in decisions so far, and how people can get involved in future decisions (including through audit, complaint, feedback and redress processes).

Suggest resources or case studies in support of this principle here.

We will also adopt the following supporting practices:

Principle 8: Institutional structures

Develop organisational capacity and embed engagement - Developing ongoing organisational capacity to meaningfully engage stakeholders around AI procurement and contracting requires an intentional approach.

Outside of individual procurement processes, create mandates, roles, and budgets to fund empowered multidisciplinary and cross-functional teams to make stakeholder engagement systematic and mainstreamed—not optional or occasional.

Standards

Supporting resources

  • We have established ongoing support for participatory practice in AI procurement.

  • All our technology procurement processes can demonstrate how affected stakeholders are involved.

Suggest resources or case studies in support of this principle here.

Principle 9: Market development

Diversify providers and partners -  Public procurement has a central and stewardship role to play in shaping the AI market, and supporting a diverse range of innovative suppliers:  beyond the default of big tech.

Procurement opportunities should be open to smaller, local, mission-aligned organisations with expertise in ethics, social justice, and sustainability whenever possible.

Standards

Supporting resources

  • We ensure procurement opportunities are accessible to small firms.
  • We actively engage with the market to describe the solutions we want to see over the long-term.

Suggest resources or case studies in support of this principle here.

Principle 10: Evaluation & learning

Reflect on and improve stakeholder engagement over time - Embedding public voice in AI procurement is an ongoing task, with opportunities for continual  progressive improvement.

Create regular spaces for reflection, learning, as well as formal evaluation of stakeholder engagement within AI procurement processes.

Standards

Supporting resources

  • We gather feedback on the engagement process from both procurement teams and stakeholders who have been engaged.

  • We discuss and share our learning with peers in order to drive continual improvement.

Suggest resources or case studies in support of this principle here.

Acknowledgements

This work draws upon Stakeholder Engagement in Public Procurement for Artificial Intelligence: a Mission-Oriented Playbook, commissioned by ParticipationAI, as well as the Toolkit for worker voice in public sector procurement of digital and AI systems in Wales developed by Connected by Data and contributions from the following people: Thai Jungpanich, Lucia Errandonea, Tim Davies, Anna Colom, Warren Smith, Abby Lupi, Freyja van den Boom, Kristina Khutsishvili, Shazade Jameson, Tim Hughes and Michael Strange. The document does not necessarily represent the views of any of the individual contributors.