Agentic development will expose GovCon's old business model

I am an artificial intelligence convert.

Despite leading a defense technology company, up until recently I didn’t consider myself an AI evangelist. It just wasn’t a priority for me personally. That changed after I met vibe coding pioneer Gene Kim at a software development workshop for the Navy.

Gene asked me a simple question: “Have I ever thought about developing an app?” That turned into a discussion about the massive void in tools for scheduling pickleball matches, followed by a spontaneous coding session in the hotel bar that ended with a working prototype of my ideal app. 

All of that happened in two hours. I was blown away by the potential this represented: agentic software development is not just another productivity tool. It can change how software gets built, tested and delivered.

What started as a vibe coding demo for a pickleball app led to a much more profound realization –AI-enabled software development doesn’t just incrementally shift how software is developed; it completely transforms the speed and agility of delivering applications from the software factory to the tactical edge.

It’s not just writing code faster. The real value is faster prototyping, modernization and mission-driven development without abandoning architecture, review, testing, integration, security or governance.  And it doesn’t eliminate the need for smart coders – it just makes them better.

For the Defense Department, that kind of speed is not a luxury. Mission software, training tools, synthetic environments and edge applications all need faster iteration. The warfighter cannot wait for the next release cycle.

Agentic development will force defense contractors to stop selling software delivery as labor throughput and start competing on mission speed, governance and outcomes. Contractors that do not change their operating and pricing models will be exposed and fall behind.

The real challenge is organizational

Agentic development changes who gets to participate. For years, software development has created a barrier between the people who understand the mission and the people who build the tool. That divide costs time, context and slows speed to capability.

AI narrows that gap and makes the process more inclusive. Business development teams, mission experts, operators and customer-facing teams can now help shape early prototypes and workflows directly. They do not need to become developers to make meaningful contributions.

That is a major advantage in the government market, where the best solutions combine technical execution with deep operational understanding. But it will not happen by accident. The winners will be the ones that drive adoption across engineering, growth, leadership, delivery and mission teams.

If software creation accelerates but compliance, testing, procurement and internal coordination do not, the bottleneck simply moves. Contractors will still disappoint customers, just with better demos. The old silos have to break down. Faster software delivery only matters if the entire organization can move at speed.

The business model must change too

The uncomfortable truth is that much of the federal software market is still organized around the economics of effort, not the economics of outcomes. Agentic development will challenge that.

Federal contractors need a new business mindset. Faster delivery will put pressure on time-and-materials models, labor categories and pricing assumptions built around older ideas of how long software should take to build.

That is good for customers. It is harder for contractors whose revenue models depend on legacy delivery timelines. But pretending that tension does not exist will not make it go away. Federal contractors must redesign pricing, capture and delivery strategies around that reality.

Agentic development has changed the basis of competition. Firms that can prototype faster, bring mission experts into the development process earlier, align engineering with growth and rethink pricing will have a real advantage.

Those that treat AI as a side experiment will fall behind.

The future of defense software will not belong to the contractor with the most people billing against the problem. It will belong to the contractor that can turn mission context into secure, tested and deployable capability the fastest.

Agentic development makes that possible. It also makes the old excuses harder to defend.

------

Matt Jones is CEO of Sigma Defense.