The Pentagon’s $1.5 trillion FY 2027 budget request is a sign of how quickly demand is rising for AI capabilities, along with ways to defend against them. The request includes $54.4 billion for autonomous warfare, up from $225 million in FY 2026, with $20.6 billion of that amount going toward systems that counter unmanned threats.
AI is helping compress military technology cycles while equipping attackers and defenders with new capabilities. As we previously examined, generative AI shrinks the time cyberattackers need to exploit a security vulnerability. In this article, we cover how AI is reshaping cybersecurity, defense procurement, and the competitive landscape for defense primes and neoprimes, powered by insights from AlphaSense’s SuperAnalyst.
AI Is Widening the Attack Surface
Every AI-driven system procured by the Pentagon adds code that bad actors can exploit. An estimated 13% of code written by AI carries security vulnerabilities, broker research shows.
Those risks multiply as the military adopts autonomous systems at fleet scale. Sell-side research indicates that the U.S. Air Force plans to field more than 150 combat-capable Collaborative Combat Aircraft by year-end 2030 — and more than 1,000 over the life of the program — with each one running autonomous software.
Cybersecurity changes everything. More and more, you have electronics in sensors, we have connected sensors, connected weapons, and a hacker can mess up everything, can mess up communication, can mess up imaging, can mess up fighters.
AI-Native Neoprimes Compete on Speed and Cost
Military technology cycles have shrunk from months or years down to weeks, making it essential that software-defined hardware is built to be adapted or upgraded, according to broker research in AlphaSense. At the same time, government buyers are prioritizing affordability and shorter turnaround times while challenging bidders to use more commercial technology. A recent Defense Department memo set fixed-price contracts as the default, limiting cost-based contracts to basic research and major system development.
Nimble, tech-first “neoprimes” like Anduril Industries, Shield AI, and Skydio are considered well positioned in this environment. Whereas legacy primes like Lockheed Martin, RTX, and Textron use cost-plus contract structures for stability and predictability, AI-native challengers compete on cost and speed.
As a result, neoprimes are already in direct competition with legacy primes for major bids. The U.S. Army awarded prototype agreements for its next-generation precision strike missile to five companies: Anduril, Castelion, Lockheed Martin, Mach Industries, and RTX.
We typically used to speak of the five top prime OEMs like Boeing, Lockheed, Northrop, and G Atomics. Now, we talk about five primes and one neoprime, which is Anduril.
Legacy Defense Primes Step Up AI Usage
As buyers have grown more receptive to flexible contract structures, Booz Allen and SAIC have built AI tools to turn cost-plus work into fixed-price bids. Lockheed Martin signed four framework agreements with the Pentagon this year to accelerate missile production, generating more than $90 billion in bookings, and its Matrix offering provides an autonomy stack similar to those of defense tech companies. Lockheed is also integrating AI into its aircraft, recently testing AI agents for live intercepts:
Our autonomous agents consumed classified infrared search and track feeds and executed combat critical maneuvers in real time. This achievement marks a decisive advance toward delivering AI-augmented air dominance for the United States.
This new paradigm carries trade-offs for legacy primes. For example, fixed-price contracts shift responsibility for cost overruns from the buyer to the builder. One program head at a major contractor puts the situation succinctly: “The days of cost plus are well and truly over.”
A former Lockheed Martin manager outlines the core challenge for incumbents: moving quickly without sacrificing mission assurance while they adapt to a more competitive, lower-margin environment. Failing to stay ahead of this trend will likely result in industry consolidation as legacy primes scale to stay ahead.
If primes want to keep Anduril at bay, they're going to have to partner with them. Otherwise when it comes to a competitive landscape, Anduril is just going to defeat them because they can acquire other start-ups, they can bring greater venture capital to bear than the primes.
Funding and Oversight Remain Key Constraints
The Pentagon’s initial budget request overstates what contractors are likely to receive. Most panelists at a recent industry event projected the final budget to come in at around $1.1 trillion base funding, well short of the $1.5 trillion headline figure even after accounting for reconciliation or supplemental funding.
Drone programs face the most scrutiny, with lawmakers reluctant to commit significant cash upfront before seeing clearer procurement plans. Systems with demonstrated strong performance in the field are considered best positioned to keep their budgets, according to sell-side research in AlphaSense.
Meanwhile, government oversight of AI vendors is tightening. In July, the White House announced the “Gold Eagle” initiative aiming to use AI for threat detection and remediation, followed by the launch of a new federal AI task force.
AI vendors have started setting their own guardrails: In September, OpenAI, Anthropic, Google, Meta, Nvidia, and others signed a voluntary accord calling for stronger internal controls, independent external audits, and a board-level review of safety findings.
Looking Ahead
For defense contractors, AI awareness and capability is becoming a condition of doing business. One contractor’s 10-K warns that failing to comply with the Pentagon’s Cybersecurity Maturity Model Certification requirements could restrict its ability to bid for, win, and perform on Defense Department contracts. The firms that can combine speed and scale to stay ahead of the technology are considered best positioned in this environment:
I feel that AI makes both cyber attacks and cyber defense significantly more capable, but the advantage will depend increasingly on who can execute this workflow at scale and at speed.
Eliminate Blind Spots With Always-On Intelligence
As AI presents in more real-world defense deployments, staying ahead means tracking how budgets, procurement rules, and competitive positioning are shifting in real time across earnings calls, broker research, and expert interviews.
Maintaining that kind of coverage manually requires more time and resources than most teams can afford. This is the issue AlphaSense’s SuperAnalyst was designed for: SuperAnalyst is an always-on AI agent that plans, monitors, and executes entire workflows on your behalf, complete with final, boardroom-ready deliverables. SuperAnalyst is grounded in AlphaSense’s industry-leading library of premium, proprietary content, meaning you can trust that every finding is grounded in a verifiable source.
Move from research to execution around the clock with SuperAnalyst, and start your free 2-week trial of AlphaSense today.
