Title

AI in public procurement: From pilots to scalable impact

Section
Insight
Summary

Author: Gil Gorev
Contributors: Kate Powell, Fabian Moreira-Phillips, Amy Bowyer, Jack Jeffries, and Jose Oliveira

As public sector procurement and commercial teams face intense pressure to deliver more with less, artificial intelligence (AI) offers a compelling way to help square this circle. Across Government, we are seeing a flurry of pilot initiatives underway – yet many organisations are grappling with how to embed AI responsibly, effectively, and at scale, within the realities of public sector governance, budgets, and structures.

In many cases, the challenge is an absence of a clear starting point or a strong foundation. Moving beyond pilots need not mean betting on unproven technologies or upending existing systems. Instead, it calls for pragmatic, outcome-focused adoption strategies anchored in real operational needs.

This article shares practical AI use cases in public procurement, common barriers holding back organisations from progressing with confidence, and steps leaders can take to move from AI experimentation to impact.
 

  • Smart tender creation – AI drafts tender documents using past procurement and market data, aligning with policy goals. Free-text AI agents can help non-procurement professionals build tender documents by offering guidance and prompts.

  • AI bid evaluation – Algorithms assess bids against set criteria, accelerating award decisions and reducing bias.

  • Procurement intake triage – AI agents or chatbots can capture and categorise internal procurement requests, apply business rules, and route them appropriately, freeing up procurement staff and improving service quality.
  • Contract summarisation and clause comparison – AI can rapidly summarise lengthy contracts and highlight deviations from standard clauses, supporting faster legal review and stakeholder understanding.

  • Performance benchmarking – AI tools can benchmark contract KPIs against similar contracts or external standards, highlighting underperformance and prompting renegotiation.
  • Market intelligence and supplier discovery – AI surfaces new suppliers, including SMEs and social enterprises, from unstructured sources, while flagging risks such as ESG issues, financial health, and adverse media to support longlisting.

  • Supplier risk monitoring – AI scans financial, ESG, and geopolitical data to flag supplier risks.

  • Bias auditing in decision-making – Algorithms can flag potential bias in procurement decisions (e.g. geographic, demographic, vendor type), supporting fair and inclusive procurement practices.