An 'Expensive Failure' – Why Smart AI Procurement is Crucial

As artificial intelligence (AI) becomes embedded across public sector systems worldwide, some agencies are procuring it without adequate research or sound judgement.

That's according to new report from the Open Contracting Partnership, which warns that governments must rethink how they buy and deploy AI or risk wasting money and eroding public trust.

Buying AI: Tools and Tips for Public Procurement offers practical guidance to help leaders make smarter choices on AI procurement and use.

Failing to deliver

OCP is an independent, non-profit public charity, working across more than 50 countries around the world. It attempts to ensure public money is spent fairly and effectively on public contracts, and with transparency to show where money is going.

The OCP is also committed to the transformation of procurement into a digital, data-driven government service that can inspire innovation.

Its new report serves as a practical guide for procurement officials to source and implement AI responsibly. With AI still treated as a catch-all solution, many buyers misuse it or fail to realise its potential. Based on interviews and workshops with more than 50 public practitioners, the research identifies common barriers and sets out where practice can improve.

OCP discovers governmental bodies are eager to adopt generative and agentic AI, but are not doing enough planning or research into the tools. Instead, they are buying AI tools without the in-house capacity to accurately integrate them or are unable to accurately challenge vendors' claims. 

As a result, many bodies are wasting public money on tools they cannot use or do not offer what they need.

“Governments are racing to adopt AI, but too often they’re doing it without the preparation and guardrails to make those investments succeed,” says Kaye Sklar, Senior Program Manager for Content and Insights at OCP, and one of the report’s authors.

“Public agencies are under pressure to ‘get into AI,’ but without strong procurement practices, they risk buying expensive tools that don’t work, don’t get used, or even undermine public confidence.”

AI procurement success stories

By setting clear AI adoption strategies that weigh opportunities and risk, governments can plan to boost operational efficiency while managing exposure. This gives agencies confidence that AI programmes will deliver and comply with legislation.

While many bodies still misuse AI or underinvest in research and preparation, others show it can work. Collaborative, flexible procurement is helping those governments adopt AI more effectively:

• The US: The Centres for Medicare & Medicaid Services (CMS) partnered with Skyward IT Solutions to procure AI and machine-learning services. This modernised operations and showed that outcome-driven contracting can accelerate innovation at the same time as saving money. The services include a chatbot which reduced ticket creation time and a procurement analysis tool which is saving CMS approximately US$3m every month.
• Australia: The Digital Transformation Agency (DTA) was able to create a robust framework for efficient and safe AI in government, which includes policies, technical standards and model contract clauses. It integrated the guidance directly into the government's digital marketplace, meaning buyers can be assured of high-quality practices while sourcing technology. The polices are presented in real time, meaning they are regularly updated to represent ongoing developments. However, DTA acknowledges that delays are caused by a lack of collaboration, for example when agencies bring in procurement teams too late to the planning. 
• Chile: The country's national digital agency created an internal “sandbox” to test and verify vendor claims before full-scale deployment.

Together, these cases show that when governments pair flexibility with accountability, AI can be a force for measurable public value.

“Poor AI procurement can cost governments millions, not just in failed projects, but in lost public trust,” comments Kathrin Frauscher, OCP’s Deputy Director. 

“Done right, though, it can modernise services, boost efficiency and prove governments can use emerging technology responsibly. This guide gives public institutions a practical roadmap to do exactly that, turning AI’s potential into measurable results for the people they serve.”

Working in collaboration

The Buying AI guide discusses the steps needed to take by governments to improve procurement readiness and AI integration. The successful implementation of AI requires a whole procurement team to navigate the challenges and accurately use AI to increase efficiency.

OCP states the key roles in this process are:

• Procurement officer - should articulate the scope and strategy of procurement and provide oversight during contract implementation
• Project team leader/buyer - facilitate overall discovery process to ensure smooth collaboration and progress
• IT/data science expert - provide guidance around responsible AI use, limitations and capabilities, as well as security requirements
• Legal officer - help navigate regulations and legality around data, liability, intellectual property and AI use cases

OCP states that they key to successful AI implementation relies upon clear planning, welcome collaboration and agility through the market.