How AI is Transforming Federal Capture: Realities, Myths and Red Lines

This month, Senior Consultant Ryan Hay hosted a deep dive into a topic dominating every corner of the public sector: artificial intelligence. To parse through the noise and identify how AI is practically moving the needle in federal business development, Catalyst welcomed back guest speaker Robin Zickgraf, Chief of Integrated Marketing turned private sector pursuit strategist at Arctic IT Government Solutions (a small 8(a) Alaska Native Corporation and Microsoft partner).

With 21 years of experience at the General Services Administration (GSA), Robin brought an invaluable perspective on bridging government regulations with modern private-sector sales workflows. Alongside insights from Ryan Hay, the discussion yielded highly tactical advice on integrating AI into a disciplined capture process.

1. The State of AI in the Public Sector: Survey Results

To set the stage, the discussion highlighted insights from a recent organizational survey revealing mixed sentiments and evolving habits within public sector consulting and business operations:

• 66% are using AI in limited ways, primarily leveraging it as a daily assistant.

• 41.7% believe AI will have the biggest long-term impact on compliance and risk.

• 50% cite the lack of clear industry regulation as their chief concern.

Furthermore, when asked where AI would strike first, business professionals reported a three-way tie among proposals/RFP responses, contract management, and sales and marketing. Interestingly, respondents were perfectly split on whether AI presents a net risk or net opportunity, with 41.7% choosing "mostly opportunity" and another 41.7% viewing it as an "equal mix of risk and reward".

2. Real Capability vs. Noise: What AI Can and Cannot Do

One of the core themes highlighted by Robin was that AI excels exponentially at high-volume data synthesis, taking the friction out of tedious tasks. However, it completely lacks the strategic judgment and emotional intelligence required to run a high-performing capture operation.

Robin: "The more you use it, the more you'll realize it's not that smart, it's just responding to what you tell it to do, and it'll draw incorrect assumptions, easily and freely, and it'll do it with a smile on its face, so to speak. So it doesn't make you smarter. But it does take that friction out, which is just a fancy way of saying it speeds things up."

AI models are fundamentally designed to be pleasing and sycophantic, which can mislead an over-reliant capture manager. As Michelle Warren added during the session, "It literally wants to make you happy and so it's just going to give you the answer you want to hear, so I think that's where there's a real skill in being able to prompt it, and ask the question the right way."

Crucially, there are several deeply human parts of procurement that AI cannot replicate:

• Build customer relationships: True capture and business development relies heavily on personal trust.

• Read a room: AI cannot pick up on subtle, real-time interpersonal dynamics.

• Give dynamic presentations: Delivering high-impact, adaptable pitches requires human agility.

• Know when to walk away: Strategic intuition and risk tolerance belong strictly to leadership.

Robin: "AI is weakest where our sales, business development capture work is the hardest. So that's building those customer relationships. They can't do that. Reading a room, giving presentations, knowing when to walk away."

3. Three Practical Capture Use Cases Robin Uses Daily

High-Speed Opportunity Research & Go/No-Go Evaluation

Instead of manually parsing dense, 60-page Statements of Work (SOW) to identify match criteria, capture teams can run files through an AI model configured with specific organizational thresholds (such as target agencies, NAICS codes, and rough orders of magnitude).

Robin demonstrated a live, real-time example using Claude to analyze an NIH statement of work that immediately returned a firm "No-Go" status. The AI flagged mismatched capability alignment and discovered a vital structural detail hidden in the text: "This OCI [Organizational Conflict of Interest] is a poison pill. If you do this project and you get awarded, you're excluded from follow-on work." Catching these hidden constraints saves massive amounts of time before pursuing an opportunity.

Structured Pre-Qualification Reporting

Once a pursuit is greenlit, formatting that data into CRM-ready layouts or leadership briefing templates can take hours. Customized prompts can be used to automatically extract incumbents, set-asides, teaming requirements, and technical scopes into highly organized summaries, enabling faster internal gates and better-informed leadership sign-offs.

Client & Partner Meeting Preparation

Understanding a partner or contracting officer's background before a call is a necessity that usually requires hours of desktop research. By using enterprise tools like Copilot, teams can rapidly spirit through publicly available internet data, LinkedIn profiles, and historical organizational touchpoints to establish structured prep sheets including communication style preferences, motivations, and tailored discussion points.

4. Red Lines and the Regulatory Horizon

The conversation concluded with an urgent warning regarding security and compliance. Both speakers heavily penalized the direct generation of bid responses using AI, noting that if an agency flags an RFP response as purely AI-generated, "we just throw it away."

Furthermore, GSA is actively entering the comment phase for upcoming artificial intelligence procurement regulations. Capture teams must prepare for an environment focused on strict transparency rules.

Robin: "Where I see it going is that you're going to have to define any time you've used AI to create something, you're going to have to tell what percentage of effort was done by AI, what information did it have access to, etc. So you're going to have to disclose all that, I think just like you would if it was a colleague or a co-worker."

Ultimately, the takeaway for federal consultants is simple: implement AI to streamline workflows and strip out administrative friction, but leave strategic thinking and relationship building entirely in human hands.

Robin: "AI works best as a catalyst, right? That's 100% true. It's not an autopilot. There are some things you can autopilot with AI, but capture business development, it just really isn't one of them. I think that’s an area that AI is never going to take anyone's job in, just because it requires a human touch, it requires judgment."

Manufacturers: Are you ready to benchmark your public sector strategy? Complete the specialized Catalyst Pathfinder Assessment to benchmark your active contract portfolio and discover exactly where your business stands in its procurement evolution!