The UK Government has demonstrated a clear ambition to harness AI to enhance efficiency, transparency, and value for money in public services, as evidenced by initiatives like the “Artificial Intelligence Playbook for the UK Government” (well worth a read to understand all the legal, ethical and practical issues relevant to the topic of AI in public sector) and Procurement Policy Notes (PPNs) such as PPN 02/24 and its updated iteration, PPN 017. However, the integration of AI into public procurement raises complex challenges that require careful navigation under existing and emerging legal frameworks, including the Procurement Act 2023. Below, I outline the key issues and their implications.
Efficiency and innovation
AI promises significant efficiencies in public procurement, a process often criticised for its complexity and resource demands. The AI Playbook highlights how AI can streamline tasks such as data analysis, supplier selection, and contract management, potentially reducing procurement timelines and costs. For instance, machine learning can analyse vast datasets to identify spending patterns, while generative AI tools can assist in drafting tender documents. The UK’s “National AI Strategy” (2021) and the subsequent “AI Opportunities Action Plan” (January 2025) underscore this potential, advocating for AI to bolster resilience and productivity across public services.
From a legal perspective, this aligns with the Procurement Act 2023’s emphasis on simplifying processes and maximising value for money. Dynamic Purchasing Systems (DPS), as noted in the AI Playbook, offer flexibility for procuring AI solutions, allowing new suppliers to enter the market and fostering innovation—a critical advantage in an evolving technological landscape.
Transparency and accountability
The use of AI by suppliers and public authorities introduces risks of opacity, which procurement law seeks to counter through transparency obligations. PPN 02/24 (March 2024) and its updated version, PPN 017 (February 2025), address this by encouraging suppliers to disclose AI use in tender preparation and service delivery. These notes provide optional disclosure questions (e.g. Annex B of PPN 017) to identify AI involvement, ensuring contracting authorities can assess its impact without scoring it in evaluations—a pragmatic balance between transparency and fairness.
However, these measures are advisory for central government bodies and optional for others, raising questions about consistency across the public sector. The AI Playbook complements this by urging alignment with ethical principles like transparency and accountability, but enforceable mechanisms remain limited. Future policy may need to mandate AI disclosure under the Procurement Act 2023 to ensure compliance with public law principles of fairness and openness.
Ethical and legal risks
AI’s reliance on data poses significant legal challenges, particularly under the General Data Protection Regulation (GDPR) and the Data Protection Act 2018. PPN 02/24 warns against using confidential tender documents to train AI models, a practice that could breach data protection laws and undermine procurement integrity. The AI Playbook reinforces this by stressing responsible data use and privacy protections in its “Using AI Safely and Responsibly” section, advising public bodies to seek legal advice on data-related risks.
Bias in AI systems is another concern. If AI tools used in procurement (e.g. for bid evaluation) reflect algorithmic biases, they could violate equality duties under the Equality Act 2010 or skew outcomes, contravening the Procurement Act’s non-discrimination principles. The AI Playbook acknowledges this, recommending oversight mechanisms to scrutinise AI decisions, but practical implementation lags. Robust due diligence, as suggested in PPN 017—such as site visits or clarification questions—will be essential to mitigate these risks.
Supplier dynamics
The integration of AI into procurement could reshape supplier dynamics. The AI Playbook and AI Opportunities Action Plan advocate supporting domestic AI startups through procurement, yet there’s a risk of entrenching market leaders. If AI systems improve through contract performance data, incumbent suppliers could gain an unfair advantage in future tenders, challenging competition principles under procurement law.
Balancing innovation with fair access—particularly for SMEs and non-specialist firms—requires careful tender design. The AI Playbook suggests leveraging frameworks like the Crown Commercial Service’s AI DPS, but regulators must ensure these do not inadvertently favour large tech firms.
Security concerns
For procurements involving national security, PPN 017 advises collaboration with Information Assurance teams to implement risk mitigations. The AI Playbook echoes this by highlighting governance and security considerations, particularly for high-risk AI systems as defined under the forthcoming EU AI Act (which, though not directly applicable, influences UK policy). Public bodies must reconcile this with the Procurement Act 2023’s flexibility, ensuring security does not stifle innovation.
Capacity and skills gap
A persistent issue is the public sector’s capacity to procure and manage AI effectively. The AI Playbook calls for upskilling teams in technical and ethical AI competencies, yet many authorities lack the expertise to assess AI tenders or monitor performance. This gap, noted in the AI Opportunities Action Plan, risks over-reliance on suppliers, blurring accountability lines—a concern under public law principles of responsibility.
Policy evolution
The future of AI in UK public procurement hinges on evolving policy frameworks. The Procurement Act 2023, provides a foundation for flexibility and innovation, but AI-specific regulations may be needed. The AI Playbook is a living document, promising updates as AI matures. Integrating these with PPN guidance could create a cohesive strategy.
The government’s pro-innovation rhetoric (especially given increased deregulation in the US) must be tempered by robust safeguards to protect fundamental rights and due process—areas where the UK risks lagging behind EU counterparts if it prioritises speed over scrutiny. A potential UK AI Act, building on the risk-based approach of the EU model, could address this, ensuring high-risk AI in procurement meets stringent standards.
AI offers a transformative future for UK public procurement, promising efficiency and innovation as outlined in the AI Playbook and supported by PPNs 02/24 and 017. Yet, its success depends on addressing transparency, ethical risks, market fairness, security, and capacity challenges within the Procurement Act 2023 framework. I advocate for a balanced approach: leveraging AI’s potential while embedding legal and ethical guardrails to uphold public trust. The UK stands at a crossroads—get this right, and we can lead globally; falter, and we risk undermining the very principles procurement law seeks to protect.
