AI Procurement in Government: Striking a Balance
Kathrin Frauscher, deputy executive director at the Open Contracting Partnership, emphasizes that effective procurement practices are crucial for successful artificial intelligence (AI) adoption in government. In a discussion with Terry Gerton, Frauscher noted that flexible and responsible procurement could bridge the gap between AI’s potential and its real-world applications, ensuring it delivers value to both government employees and the citizens they serve.
Frauscher advocates for what she describes as “flexible” procurement practices, which allow for swift adjustments based on the specific AI tools needed. This approach contrasts with traditional procurement processes, which are often mired in bureaucracy. “Speed matters because procurement has the reputation, but also indeed often is rather slow and bureaucratic,” she stated, underlining the importance of maintaining agility in a fast-moving technological landscape.
Interestingly, research conducted by the Open Contracting Partnership revealed a preference among government agencies for “off-the-shelf” AI solutions rather than custom-built applications. This finding highlights a shift in procurement strategies, as many agencies are now relying on existing platforms rather than lengthy, bespoke requests for proposals. This transition necessitates a reevaluation of the procurement framework to ensure it aligns with current technological realities.
Balancing caution and urgency remains a challenge for procurement leaders, especially given the inherent risks associated with AI, such as bias and errors in decision-making. Frauscher stressed the importance of finding a middle ground: “You can’t move too slow so that you’re missing out on things, but you can’t move too fast so that you’re not ready.” Establishing a governance framework is essential for managing this balance, as it provides clear guidelines for the responsible deployment of AI technologies.
Governments must also invest in training their personnel on these frameworks. Frauscher pointed out that having a governance structure is insufficient if employees do not understand how to implement it effectively. “So don’t just have it on a piece of paper, but make sure that people know how to use it,” she advised.
Although the discourse often surrounds the procurement of AI tools, Frauscher highlighted the potential for AI to enhance the procurement process itself. Current applications primarily assist with tasks such as drafting requests for proposals and streamlining workflows, which may appear mundane but are crucial for operational efficiency. “Boring things are very important too,” she remarked, signifying that even basic applications of AI can yield significant benefits.
Coordination among procurement, IT, and project teams is crucial for successful AI acquisitions. Frauscher described procurement as a “team sport” and emphasized the need for collaboration from the outset. This collaborative approach ensures that everyone involved is “AI ready,” equipped with the requisite knowledge to make informed decisions about technology purchases.
The General Services Administration’s initiatives to centralize AI procurement are part of a broader global trend aimed at enhancing efficiency and standardization. While centralized purchasing offers advantages, Frauscher warned that implementation must remain adaptable to the specific needs of individual agencies. “Each agency has the capacity to use AI well,” she noted, emphasizing the importance of tailored solutions.
In light of the challenges posed by vendor exaggeration regarding AI capabilities, procurement teams must adopt a critical approach when evaluating potential products. Frauscher recommends asking vendors for demonstrable success stories and advocating for prototypes to validate claims, which can help cut through the hype and ensure tangible value.
The Open Contracting Partnership has developed a comprehensive guide to assist procurement officials in navigating the complexities of AI purchasing. The guide addresses foundational knowledge about AI and outlines collaboration strategies for procurement and project teams. It emphasizes practical scenarios to illustrate various purchasing pathways, helping officials to make informed decisions that align with their specific goals.
As AI technology continues to evolve rapidly, Frauscher envisions a future where procurement processes are agile enough to adapt. Reflecting on the potential landscape five years ahead, she noted, “AI is changing and evolving at a speed that we haven’t seen that often before.” This adaptability will be essential for governmental organizations to meet emerging challenges and harness the full potential of AI.
For more insights on AI procurement and its implications for government, visit the Open Contracting Partnership.
To learn more about government procurement standards and practices, check the General Services Administration.
For further reading on artificial intelligence’s impact on various sectors, consider resources from MIT.
See also
Google Launches Gemini Deep Research Agent with Enhanced Web Search and Reporting Accuracy
AI Transforms Military Supply Chains: Enhancing Resilience and Readiness Amid Fragility
Google’s Gemini 2.5 TTS Revolutionizes AI Voices with Human-Like Emotion and Precision
Anthropic Plans IPO Amid $183 Billion Valuation and Strong Market Sentiment
Nvidia’s H100 Chip Powers First Space AI Training, Aiming for 5GW Data Center by 2026
