8 Key Themes from NDIA’s Emerging Technologies for Defense Conference 2025
By: ETI Research Team
Over three days in August 2025, the NDIA Emerging Technologies for Defense Conference brought together senior leaders from government, industry, and academia for pressing conversations on challenges and opportunities facing the defense community.
In addition to breakout rooms, technology demonstrations, a global hackathon, and the exhibition floor, the conference also featured a robust set of keynote sessions. Across the dialogue, eight key themes stood out:
1. “Move with Speed and Purpose”
The main theme for the 2025 NDIA Emerging Technologies Conference was set at the opening by Gen. Dan Caine, USAF, Chairman of the Joint Chiefs of Staff: “we need decision makers who clear the path remove obstacles, so that the best ideas move with speed and purpose… Our competitors are not waiting. They’re building faster, integrating more aggressively, and accelerating from concept to fielded capability at scale.”
Speakers throughout the conference provided perspective on the practical bottlenecks that enable speed. Darren Kimura, CEO of AI Squared, highlighted one that every founder feels: “Right now, it takes about 240 days to get a person through a clearance… that impacts speed to delivery.” Rear Adm. Cameron Chen, USN, Deputy Director, J3, AFRICOM, flagged cross-service friction that saps momentum: “It’s certified on an Air Force ATO, but then we have to redo the whole process to get the Navy to use the almost exact same thing. That definitely slows things down.” Stuart Wagner, CDAO at the Navy, focused on decision speed: “The speed to product-market fit is slow in the DoD… Getting a ‘no’ is better than stringing you along with a fake ‘yes.’” For industry, speed provides rapid market feedback, so teams can pivot or scale.
USW(R&E) Emil Michael described how DOW is speeding up by opening the front doors to innovators, using the flexible authorities it already has, and giving companies a fast yes or fast no. “We don’t want the Department to be a black hole where good ideas go to die… we’re opening front doors,” he said, pointing to DIU, CDAO, DARPA, and R&E test venues like the recent T-REX event where 103 drone companies trialed in a realistic environment and received immediate feedback.
The NDIA Global Hackathon (41 teams) brought that mindset onto the conference floor. Nick Lanham, Director, Enterprise Platforms and Data, KAIROS Inc., noted that the Hackathon resulted in demos “built in the last 72 to 48 hours,” and Maj. Culley Horne explained why it worked: hackathons “take the engineer and the operator and put them very close together,” so prototypes built against problems transition faster.
2. Open Architectures with Mission Integration
Speakers argued that the way to get both speed and competition is to integrate systems within an environment of interoperability. Former SAIC CEO Toni Townes-Whitley framed the objective end-state for mission integration as “full real time, secure interop,” enabled by “open architecture…plug and play of emerging tech…we can’t build vertical systems as we used to. We now have to build in a more horizontal way.” But interoperability isn’t just an acquisition approach: it’s a coalition imperative, as underscored by Gen. Alexus Grynkewich, USAF, SACEUR, stating, “Stitching ourselves together is going to be essential.” Rear Adm. Chen agreed, “To counter China, it has to be a partnership approach with allies for a common picture.”
Cloud providers and data platforms echoed the same pattern. Dominic Delmolino, Vice President at AWS, described a coalition model built on “building blocks – secure primitives – for partners to assemble mission-specific solutions across multiple classification levels.” Dan Tadross, Head of Public Sector at Scale AI, underlined the human reality: “Anyone who tells you they can do it all is telling something a little less than the truth. The way we scale is partnerships – and policy, business, and technology levers that foster the ecosystem.”
3. New Technology Pathways
Gen. Caine noted the opportunity and challenge posed by a new generation of commercial technologies: “The barrier to entry for disruptive technology is low, but the barrier to long term government partnerships is still too high.” The services explained their tactics for keeping pace: for the Army, Acting ASA(ALT) Jesse Tolleson discussed tactics “to focus on using an ‘adopt first, modify second, develop last’ type model,” and pairing that with “characteristic of the need” statements that frame problems broadly, then tighten via experimentation to ensure programs don’t spec themselves out of viable solutions. For the Navy, acting ASN(RDA) Jason Potter pointed to a functioning pipeline with DIU and Navy channels: “We’ve gone into production on both…OTAs as well as FAR-based contracts,” including “nearly $400 million worth of unmanned surface vessels.”
One bottleneck discussed was access. Darren Kimura described one successful tactic: “How do we meet you? In the DoD world it’s really hard…We’ve seen ‘speed-dating’ events where startups sit with mission owners who have an actual problem – that works.” USW(R&E) Michael is trying to make those on-ramps visible and paced: SBIR for ideas, APFIT funding for $20-$50M expansion, Office of Strategic Capital (OSC) for scaling, DIU/CDAO for “fast yes, and fast no.” He called it “opening all the front doors, as opposed to…side doors.”
4. Need for Surge Capacity
An important theme from government speakers was the urgency to create surge capacity for production. USD(R&E) Michael argued the “need to surge is part of deterrence,” with USW(A&S) Michael Duffey agreeing that “future conflict will be decided as much on the factory floor as the battlefield.” From the COCOM perspective, Gen. Grynkewich (SACEUR) urged U.S.-EU industrial teaming, and even plugging into Ukraine’s wartime industrial base, to learn and produce at the speed the modern battlefield demands.
In industry, speakers highlighted the challenges and opportunities to increase surge capacity. Toni Townes-Whitley described rebuilding the MK-48 Heavyweight Torpedo line with digital engineering, only to face a “22-month lead time…single points of failure in the supply chain,” gaps in testing and procurement.
To finance scale-up gap, Barclays’ Matt Spence argued that the bottleneck isn’t seed money; it’s the hundreds of millions, or sometime billions of dollars that are needed to scale hardware and manufacturing. However, the private capital panel noted that both private and public markets continue to drive large amounts of private capital toward defense technology companies.
5. Adaptability is a Key Performance Parameter
Gen. Grynkewich gave the conference a memorable call to action: “If there was a KPP for future capabilities…that KPP is adaptability.” He described Ukrainian units watching drones work on Monday, fail by Wednesday under electronic warfare, and be effective again the next Monday, only because industry partners iterate with them in near real-time: “If you don’t have the ability to rapidly adapt…you’re going to lose.”
The Army’s Jesse Tolleson is institutionalizing this concept with “characteristic of the need” constructs and rapid contracts that convert fielded ideas into programs without waiting years. Additionally, according to Adm. Christopher Grady, USN, Vice Chairman of the Joint Chiefs of Staff, the right people have to be placed in the right places: “Let’s bring the operator and the engineer together sooner, so that they can understand the key operational problem we’re trying to solve.”
Adaptability must also be built into the research and engineering process. From Boeing’s Phantom Works, General Manager Colin Miller noted: “Instrument even things that you don’t think are central…If you can’t measure it, you didn’t learn it.” Rear Adm. Chen noted that “The programs that go best have all three represented – the warfighter at the tactical edge, government expertise, and industry – getting feedback in realistic exercises.” Adm. Grady wants to “triple the capacity that the joint force has for the chairman” in modeling & simulation. Force design must be evidence-based, and evidence comes from mission-level analysis and rehearsal. On the industry side, Colin Miller described Boeing’s Virtual Warfare Center with five networked sites, “over 1,000 events,” and utilization “off the charts” – doubling engagements year-over-year – so the team can “do mission engineering… [with] the whole joint coalition family of systems in [the] customer’s threat environment.”
6. Autonomy with affordable mass: the new high-low mix
Gen. Grynkewich reframed the Cold War’s “high-low mix” for today: some high-end, sometimes manned, exquisite capabilities to win air superiority, plus “affordable mass” at the low end – largely unmanned – for volume, persistence, and attritable effects. “A large part of the future…would be unmanned and digital,” but “those one-way attack drones are not going to gain air superiority.”
USW(R&E) Michael emphasized “mass attritable weaponry” as a pillar of modern deterrence. Adm. Grady focused on survivable cognition: “I want AI at the edge, something small and fast…such that when I get cut off from the rest of the cloud…they can continue to fight and apply AI.” He also tied autonomy to day-to-day readiness: “Why is it that every time the ship pulls in, I don’t put my robotic hull cleaner on board… That’s a faster ship, more fuel efficient…Everything works better.”
Autonomy isn’t theoretical anymore; several speakers showed how it is working and scaling. Colin Miller described a cockpit where sensor fusion tells the crew what to do (and then why), and autonomy where “one button push” can execute large portions of the mission under commander’s intent. Boeing’s Ghost Bat has already flown in formation, integrated third-party C2, and is now beginning to work with live ordnance.
Critically, Gen. Grynkewich warned against tech for tech’s sake: “You may have a really neat technology…what’s compelling is a concept of operations for how you’re going to gain military effects with it.”
7. Space at software speed, and “as-a-service” signals that stick
Maj. Gen. Stephen Purdy, Jr., USAF, Senior Acquisition Executive at the USSF, laid out a vision for space acquisition that operates at software speed. Already, the USSF is accelerating satellite development, from a previous tempo of 10+ years to 24-36 months as the norm – and Purdy expects in the future “to see those satellites built in a month,” then “a couple of days.” Maj. Gen. Purdy described how the USSF has accomplished this with contracting tranches, multi-vendor awards, and recurring competitions, so that learning compounds and no single vendor “owns” a capability.
The Space Force is also leaning into as-a-service acquisition models to create persistent demand signals that new entrants (and their investors) can rely on. Purdy cited launch as a service: NSSL Phase 3, Lane 1 laid out “over 40 missions” which could be bid upon, giving private investors clear demand signal instead of a “one-and-done" award. On the network side, the Space Force is buying “large amounts ofbandwidth” through the Commercial Satellite Communications Office (CSCO) and standing up data as a service for SSA, GPS, and “the jamming environment.”
8. Operational AI
USD(R&E) Michael defined credible AI adoption with clear metrics. Today DoD has “fewer than 50,000 [AI] users per month… out of 3 million,” with the goal of measuring usage and productivity and converting “road maps” into “sprints.” Navy CDAO Stuart Wagner described an AI-enabled “bits-to-effects” pipeline: “LLMs are accelerating the ‘bits-to-effect’ cycle – you don’t even need to write code. The challenge is we’re siloing data in chats; we need LLMs hooked to enterprise data so we learn at the enterprise.” Rear Adm. Chen emphasized the need for workforce upskilling: “How do you train and certify someone to use AI and get an expected outcome?”
