People discussing marketing strategy and performance data during a meeting in a modern workspace

Right-Sizing AI: A Small GovCon Isn't a Small Enterprise

Summary

AI strategies for small GovCons do not require an enterprise-sized playbook. The right level of structure depends on the work, data, and contractual requirements involved. A focused adoption strategy can help small firms manage risk while putting AI to work where it matters most.

AI guidance for government contractors can often assume resources that smaller firms simply don't have: a dedicated compliance function, a data team, or someone responsible for developing an AI strategy. That can leave small GovCons trying to apply an enterprise AI playbook to a business with a different headcount, budget and risk profile.

Small GovCons can take a more focused approach. The size of the team alone doesn't determine how much structure AI requires. The work AI is supporting, the data it touches, and the contractual or compliance requirements involved matter, too. That means a five-person firm may need tighter controls around one use case while taking a much lighter approach to another. The questions below came directly from small GovCons who attended our State of AI in GovCon webinar, and they reflect some of the practical decisions smaller firms are navigating today.

When Should We Actually Build an AI Workflow?

Question: "At what point should a government contractor look to implement AI workflow recommendations? Should a 'one-man shop' do this? Two folks? Three? When do you stop stretching and build the foundation?"

Headcount can be part of the equation, but it shouldn't be the only consideration. The type of work you perform, the requirements tied to your contracts and certifications, and the data an AI-enabled workflow would touch can all affect the level of oversight needed. A small contractor working on a cost-reimbursable DoD contract, for example, may face very different considerations than a larger firm performing lower-risk commercial-item work.

As you evaluate an AI use case, start with a few practical questions:

  • Does the work involve certified cost or pricing data?
  • Are there contract or certification requirements that could affect how AI is used?
  • Would the AI-enabled workflow access data associated with a cost-reimbursable contract?

If any of those apply, it may be time to take a closer look at how the tool, data and workflow are governed before moving forward.

Should We Fix Our Operations Before We Touch AI?

Question: "Most small and mid-sized GovCons have bootstrapped governance and PM tasks. Our advice has been to fix operations processes before jumping into AI. What do you recommend?"

You don't necessarily have to choose one or the other. Waiting until every process is perfectly organized can keep AI experimentation perpetually out of reach. But adding AI to a broken process won't fix the underlying problem either.

Instead, start with a focused use case that gives you a practical reason to improve the process behind it. If you want to use AI to vet a draft proposal against past performance, for example, you first need to organize that past-performance data so it can be used effectively. You're improving the process and testing a useful AI application at the same time.

What AI Risks Are Unique to a Small Defense Contractor?

"What are the specific AI risks and mitigation actions applicable to small defense contractors?"

For a small defense contractor, one of the key AI risks is how sensitive data is handled when employees use AI tools. If someone enters CUI or ITAR-controlled technical data into a commercial AI tool, that information may be processed or stored in an environment that isn't appropriate for that data. Access to ITAR-controlled information can introduce additional concerns if the tool, its infrastructure or its outputs are accessible to unauthorized foreign persons.

Managing these risks starts with understanding both the data and the AI tool. Know how the information is classified, where the tool processes and stores it, how inputs may be used, and who can access the tool and its outputs. Those questions become especially important when a workflow involves CUI or ITAR-controlled technical data.

Which AI Tool is Right for the Job?

Question: "Do you recommend small business government contractors use a commercial AI platform or tools such as ChatGPT, Claude, or CoPilot?"

The right choice depends in part on the data each tool can access and the specific business need you’re trying to address. It’s less a question of which individual tool is “best” and more a question of the level of access, context, and governance each one offers. A general-purpose AI tool can be useful, but you may run into challenges once you’re trying to set up workflows that pull in the right data, especially if that data contains CUI. An AI solution built specifically for government contracting, by contrast, is often already designed around that data and the way your team actually works, and may continue to pick up context as your team uses it over time. There’s also a practical cost consideration: building and maintaining those same workflows yourself in a general-purpose tool can end up costing more, in setup time and ongoing use, than a solution that’s already built to do it.

As you evaluate your options, look closely at the data behind the tool. What sources does it use? Does it have access to the information needed for your use case? How is that information maintained and validated? Understanding where the answers come from, and how much confidence you can place in the underlying data, can be just as important as comparing individual features."

How Can a Startup Use AI to Grow?

Question: "How can AI be used in your startup to help you expand and grow, and how can it be implemented to get the greatest results?"

For a growing GovCon, the best place to start with AI will depend on where the business is feeling the most pressure. Look for focused use cases where AI can support work your team already does and where the underlying data, processes and controls are in place to use it effectively. Purpose-built solutions can be especially valuable because the AI is embedded in the context of the work, rather than requiring a small team to build that context from scratch. The goal should be to solve a meaningful business problem in a way that can scale as the company grows.

Small Can Be an Advantage with AI

Small GovCons may not have the AI resources of a large enterprise, but they also don't need to approach AI the same way. A smaller team can be selective about where AI is used in the business, focusing on specific problems where the data, workflow, and risk are well understood.

That may mean starting with one use case rather than a broad AI initiative, or choosing technology that brings AI into the systems and processes the business already relies on. The goal for a small business isn't to adopt AI everywhere. It's to put it to work where it can make a meaningful difference, then expand from there as the business and its needs evolve.

Guide

Close the AI Maturity Gap

See where AI is creating value across finance, business development, pricing, and project execution, plus the risks that can hold maturity back.

Contributors

Author

deltek company logo

Deltek

Deltek experts and industry thought leaders share their views on the latest trends, technology and issues shaping project-based businesses today.

Featured Thoughts

Dela

Article

Stop Switching Tabs: Your Favorite AI Tools Now Work With Costpoint Through MCP

This article explains how Costpoint now supports Model Context Protocol (MCP), so Dela can reach approved outside tools and the AI tools your teams already use can work with Costpoint. It shows how that cuts the time spent switching between systems, with administrators deciding which connections exist.

Deltek Product Release Information

Article

Why Finance’s Perspective Belongs in Proposal Development, Not Just After the Award

This article explains why Finance belongs in proposal development earlier, using a lesson from three proposals submitted at once and the requirements that were missed. It also shows how Costpoint Proposals helps teams bring Finance into requirements review while the response is still taking shape.

Woman looking at reports

Article

One Experience, Every Device: A Look at What's New in Costpoint

This article shows existing Costpoint customers how to get more value from ongoing improvements across the desktop experience, Costpoint Mobile, and Ask Dela. It highlights practical usability updates, customer-driven refinements, and the business outcomes supported by a more connected experience.

Aerial View of the Pentagon

Article

The Door to Government Contracting Just Got Wider

This article breaks down DoW’s memo raising the Cost Accounting Standards threshold to $100 million, covering what’s changing, what isn’t, and why cost visibility still matters for contractors either way.

The GovCon AI Gap: Insights for Building AI Maturity

Guide

The GovCon AI Gap: Insights for Building AI Maturity

AI use has become nearly universal in government contracting — but maturity hasn't yet caught up! Read this guide to learn where AI is creating value, where risks remain, and what top-performing contractors are doing differently.