How AI can aid in procurement and purchasing for cities and counties

Cities and counties, like their counterparts at the state level, are under constant pressure to deliver more with fewer resources. Artificial intelligence is a tempting lifeline, but the cost and uncertainty leave many leaders hesitant — and pilots that stall before showing real results only fuel that. That makes it all the more important to find an area where AI can show value quickly.

Procurement and purchasing are a natural starting point for AI. The work is highly structured, frequently repetitive and foundational to nearly every government function. Unlike areas such as health and human services — where guidelines for AI use are still evolving — procurement already operates within established frameworks and proven precedents. This makes it a practical proving ground where agencies can deliver measurable value quickly while keeping risks manageable.

Why Procurement, and What AI Can Do

Anyone who has worked in procurement knows the grind: reading through dense contracts, pulling together lengthy statements of work, or tracking deadlines buried in spreadsheets. AI can’t eliminate the work, but it can take the edge off by handling some of the most time-consuming tasks, such as:

• Summarizing vendor responses and policy documents for quick reference
• Creating first drafts of documents, such as requests for proposals, invitations to bid or renewal notices, to give staff a head start instead of a blank page
• Organizing questions and feedback during evaluations, so teams aren’t wading through messy email threads
• Surfacing contract clauses or tracking key dates post-award that might otherwise slip through

These are practical, here-and-now applications that tackle the rules-driven and document-heavy work of procurement officials. It’s a place where AI can make life easier for staff while they tackle important decisions.

How to Use AI in Procurement Safely and Effectively

Applying AI in procurement requires both caution and intention. The following principles can help guide a safe and effective approach.

• Start small. One way to start is to focus on a single, concrete pain point — maybe vendor-response summaries or renewal tracking — and commit to fixing that. Run a small pilot with willing early adopters and use their feedback to refine the tool and processes. Don’t forget to set clear exit criteria so the project doesn’t get stuck in pilot purgatory. Once you’ve proven value, expand deliberately
• Integrate thoughtfully. Integrate thoughtfully. Make the tools work where your people already work. AI must integrate into the systems that procurement teams rely on — including finance, contract management, archiving and vendor portals — to save time. Otherwise, staff end up copying information between platforms, double-checking entries or losing track of documents altogether. Integration is what turns AI from another experiment into a real workflow improvement
• Get the data right. Clean, structured, current information is the fuel for useful results. Build documents in stages — scope, requirements, timeline — instead of asking for an all-at-once draft, which tends to be vague. Throughout, keep humans in the loop. AI can draft, summarize, and organize, but procurement officers still need to review and sign off on any documents before they become official
• Set guardrails. Innovation shouldn’t outpace governance. Provide approved tools and training to prevent teams from bringing in unapproved shadow IT, and draw clear boundaries. For example, don’t use AI for legal or policy interpretation, vendor scoring or public-facing communications. For structure and accountability, anchor your approach on established guidance, such as the NIST AI Risk Management Framework or state policies like California’s executive order on AI

Practical First Steps for Local Governments Using AI

The good news is that cities and counties don’t have to start from scratch. Begin by speaking with agencies of a similar size or complexity. Ask them what worked, what didn’t, and what they would do differently. These peer-to-peer conversations often reveal the most practical insights.

Take advantage of cooperative purchasing agreements that already include AI solutions. Because the terms have been competitively bid, they can help shorten the procurement cycle and reduce risk.

Look for opportunities to try before you buy. Some states offer AI innovation labs or sandbox environments where agencies can test tools in a low-stakes setting. That way, staff can build familiarity and confidence without locking into a contract too quickly.

And don’t underestimate the value of professional networks. Procurement officials can compare notes and share lessons learned through associations such as the National Institute of Governmental Purchasing, the National Association of State Procurement Officials and the National Procurement Institute. Tapping into those networks can help your team avoid reinventing the wheel and dodge common missteps.

Get Procurement Right, Get AI Right

AI can’t and shouldn’t replace human judgment. But it can clear the way for procurement professionals to think strategically, shape stronger contracts, and deliver more value for their communities. By starting with procurement, cities and counties can test AI in a structured, lower-risk setting that is ripe for efficiency gains.

The formula for success is straightforward: start small, integrate into existing systems, clean up the data, and always keep humans in the loop. Done thoughtfully, AI can help procurement teams do more with less while still protecting the accountability and transparency that the public expects.

Get procurement right, and AI stops being an experiment. It becomes the launchpad for smarter adoption across government.

Stephanie Weber is industry lead for U.S. state and local government at Appian.