Interview with Ylli Bajraktari
CSDS-SWJ STRATEGY DEBRIEFS • 10/2025
Interview with Ylli Bajraktari, by Octavian Manea
This interview is part of a collaborative initiative with the Centre for Security, Diplomacy and Strategy.
Ylli Bajraktari, President and CEO of the Special Competitive Studies Project.
Before launching the Special Competitive Studies Project (SCSP), Ylli served as the Executive Director of the National Security Commission on Artificial Intelligence (NSCAI). Prior to joining the NSCAI, he served as Chief of Staff to the National Security Advisor LTG H.R. McMaster, held a variety of leadership roles for former Deputy Secretary of Defense Robert Work and served as Special Assistant to the Chairman of the Joint Chiefs of Staff, General Dempsey.
Octavian Manea: How would you define an offset strategy?
Ylli Bajraktari: Fundamentally, an offset strategy is a deliberate, long-term competitive approach to counteract an adversary’s advantages – often numerical superiority or geographic position – by leveraging asymmetric strengths. It is about changing the paradigm of competition rather than competing symmetrically. The goal is to restore deterrence by fundamentally altering how forces operate and imposing costs on the adversary.
Historically, we look at two precedents. The First Offset (1950s) was built around Eisenhower’s “New Look”, leveraging United States (US) nuclear superiority to offset the Soviet Union’s conventional dominance in Europe. The Second Offset (1970s-1980s) recognized the arrival of nuclear parity and focused on restoring conventional deterrence through the fusion of stealth, precision-guided munitions (PGMs), satellite reconnaissance and networking (the “Reconnaissance-Strike Complex”).
The Third Offset Strategy (3OS), initiated around 2014, was born from the recognition that the advantages of the Second Offset were eroding. Adversaries, having studied the American way of war, developed sophisticated Anti-Access/Area Denial (A2/AD) capabilities specifically designed to neutralize our precision battle networks.
Octavian Manea: What were the distinctive traits of the Third Offset Strategy (3OS)?
Ylli Bajraktari: The distinctive traits of 3OS focused on the next wave of technological advancements needed to sustain US power projection. The core thesis was that the fusion of Artificial Intelligence (AI), autonomy and human-machine collaboration would provide the decisive edge. It recognized that the speed of future conflict would exceed human decision-making capabilities. Key pillars included autonomous learning systems, human-machine collaborative decision-making, advanced manned-unmanned teaming (MUM-T) and network-enabled autonomous weapons.
Octavian Manea: Did the 3OS have a coherent theory of victory?
Ylli Bajraktari: Yes. The theory of victory was deterrence by denial through conventional overmatch. The goal was to credibly demonstrate that the US could identify, target and neutralize aggression, even within highly contested A2/AD environments, faster than the adversary could achieve their objectives. It aimed to prevent a fait accompli.
Octavian Manea: There are many ways to understand and define the notion of strategy. What are the specific implications of conceptualizing the 3OS as a competitive strategy?
Ylli Bajraktari: Framing 3OS as a competitive strategy means viewing it through the lens of long-term competition. It was not just about winning a specific war; it was about maintaining a durable advantage over decades. This involves making strategic investments that impose costs on adversaries, forcing them to invest in areas where we are stronger or where their investments are less effective, thereby sustaining our military-technical edge and agility.
Octavian Manea: Let us unpack the structural context that gave rise to the 3OS. There was a concern that peer competitors were on the verge of achieving parity in PGMs and theater-level battle networks. But why is parity perceived as dangerous, as something to be avoided at all costs?
Ylli Bajraktari: The broader context was the diffusion of the Second Offset. Competitors achieved parity not globally, but locally, in specific technologies crucial to US power projection: PGMs, ISR and theater battle networks.
The US way of war is fundamentally expeditionary. We fight “away games”. This requires access to forward bases, the ability to move forces across oceans and the assurance that those forces can operate effectively upon arrival. Parity in precision battle networks fundamentally threatens this model. If an adversary can see what we see and strike what we strike, they can target our critical nodes – our carriers, airbases, logistics hubs and command centers.
Why is parity dangerous? There are 3 consequences that we should have in mind.
Firstly, there is an erosion of deterrence. US foreign policy relies on extended deterrence – the promise to defend allies. Localized parity emboldens adversaries to attempt rapid aggression, believing they can achieve objectives before the US can effectively respond.
Secondly, there is the vulnerability of US assets. Parity in PGMs makes the foundation of US expeditionary warfare highly vulnerable.
Lastly, there is the reality captured by the “stability-instability paradox”: while nuclear parity might induce stability, conventional parity can induce instability. If a competitor believes they can achieve a rapid conventional victory, they are more likely to attack.
Parity was a call to action because the status quo was no longer sufficient. We needed to develop entirely new ways of fighting that negated the adversary’s investments in A2/AD.
Octavian Manea: What were the fundamental characteristics of what it is called the traditional American Way in Warfare, often referred to as the Desert Storm model? What made it obsolete?
Ylli Bajraktari: The “Desert Storm model” was characterized by four dimensions. First, through “sanctuary and buildup”: the assumption that the US could utilize secure forward bases (like Saudi Arabia) to build up an “Iron Mountain” of logistics and equipment over several months, free from enemy attack. Second, through “undisputed domain dominance”: assumed control of the air, sea, space and the electromagnetic spectrum (EMS). Third, “methodical, sequential operations”: gaining air superiority, degrading enemy C2, followed by decisive ground maneuver. Finally, via centralised Command and Control (C2): hierarchical command structures relying on superior C4ISR.
Competitors adapted specifically to counter it and targeted centers of gravity such as access and networking. Concretely, China and Russia developed A2/AD strategies to prevent the US from executing the Desert Storm playbook. They concentrated their efforts on: 1) attacking the buildup: they developed massive arsenals of precision ballistic and cruise missiles to target US bases, ports and aircraft carriers from the outset. There is no longer sanctuary; 2) contesting the domains: they invested heavily in sophisticated Integrated Air Defense Systems (IADS), counter-space capabilities and cyber/EW to deny automatic dominance; and 3) targeting the network: they focused on degrading the C4ISR architecture (the satellites and communication links) that the US relies on.
In short, the Desert Storm model assumes an uncontested buildup followed by a synchronized campaign. Today, the fight begins immediately, everywhere, and the buildup itself is the first target.
Octavian Manea: You worked very closely with Bob Work at the peak of the 3OS, so I am wondering about your take on the notion that the Deputy Secretary of Defense was emphasizing back then, especially in the context of deteriorating security on NATO’s eastern flank. You will recall that he made the case for a Multi-Domain Operational Fires Network and for ‘light infantry hunter-killer teams with anti-tank weapons and the ability to call-in precise highly lethal fires’. Today, we hear the same discussion but in an Indo-Pacific context – developing a Joint Fires Network (JFN) is literally at the forefront of the warfighting Indo-Pacific posture. To what extent is such a JFN a legacy of the 3OS? To what extent is such a multi-service JFN the answer to the needs to blunt and deny aggression in the first island chain?
Ylli Bajraktari: Deputy Secretary Work’s emphasis on “Operational Fires Networks” was central to the 3OS vision. The concept of distributed “hunter-killer teams” reflected the realization that large, centralized platforms are too vulnerable in a precision-strike environment. Survivability requires distribution, mobility and low signatures.
The JFN – I hope was inspired by the 3OS. The 3OS identified the need to connect any sensor to any shooter across all domains to achieve decision advantage. This vision is the core principle of today’s JADC2 (Joint All-Domain Command and Control) efforts.
In the Indo-Pacific, a JFN is critical for blunting aggression. To deter an invasion of Taiwan, the US must rapidly sense, understand and act. This means connecting space sensors, Marine Corps littoral units, submarines, autonomous drones and long-range bombers in real-time. A JFN enables concepts like Distributed Maritime Operations (DMO) and Expeditionary Advanced Base Operations (EABO), creating a dense web of fires.
Octavian Manea: How fungible is the idea of a Joint Fires Network from a cross-theatertheater perspective?
Ylli Bajraktari: The concept is highly fungible, but the implementation must be tailored.
In Europe, geography is compressed, and the threat is primarily land-based. The emphasis is on air defense, counter-battery fire and integrating NATO allies. Bob Work’s “hunter-killer teams” concept is tailored to this environment.
The Indo-Pacific theater is primarily maritime and aerospace, characterized by the “tyranny of distance”. The emphasis is on anti-ship fires, long-range precision strike and resilient communications.
The underlying JADC2 architecture should be the same, but the specific sensors and shooters plugged into that network will differ.
Octavian Manea: To what extent is there progress in actually developing effective new operational concepts – new ways of fighting? In the days of the 3OS there were calls for a new Assault Breaker.
Ylli Bajraktari: Progress has been significant at the conceptual level across the services:
the Army focused on Multi-Domain Operations (MDO) and the creation of Multi-Domain Task Forces (MDTFs); the Navy/USMC focused on Distributed Maritime Operations (DMO) and EABO; the Air Force developed the Agile Combat Employment (ACE) construct, moving away from large bases to dispersed locations to generate combat power under attack.
The “new Assault Breaker” (Assault Breaker II) is not a single program but the integration of JADC2, long-range fires, autonomous systems and these operational concepts. It is a system-of-systems approach designed to “break” an adversary’s attempt to seize territory, such as a large-scale amphibious assault in the Western Pacific.
The legacy of the 3OS is that it provided the intellectual foundation for the current US approach to warfare.
- 3OS provided the “Why” (the eroding advantage).
- MDO, DMO, and ACE are the “How” (the operational concepts).
- JADC2 is the “Means” (the technical architecture).
The 3OS catalyzed the shift back to great power competition, prioritized key technologies like AI and autonomy and institutionalized innovation (DIU, SCO).
Octavian Manea: Autonomy was a key component of the 3OS. How do initiatives and programs such as Replicator & Hellscape fit into the emerging US approach to warfare, and what do they reveal about new ways of fighting – particularly within the First Island Chain?
Ylli Bajraktari: Autonomy was central to 3OS as a way to offset adversary mass and operate at speed in contested environments. “Replicator” (fielding thousands of attritable autonomous systems at scale) and the desired operational effect of “Hellscape” (saturating the battlespace, particularly the Taiwan Strait, with autonomous systems) are the realization of this vision, driven by the urgent need to counter China’s advantages in mass.
They are integral of the US emerging approach to warfare by: 1) “generating asymmetric mass”: if the US cannot match China ship-for-ship, it must generate mass through large numbers of low-cost, autonomous systems; 2) “shifting the cost curve”: cheap, attritable systems impose significant costs on adversaries who must defend expensive platforms; and 3) “enabling distributed operations”: autonomous systems are key enablers for DMO and EABO, acting as persistent sensors and strike platforms where manned systems cannot operate.
Overall, there is a shift towards decentralized, autonomous and saturation-based warfare. This emphasizes swarming tactics, human-machine teaming (humans “on the loop”, not “in the loop”), and the critical importance of software-defined capabilities and rapid adaptation.
Octavian Manea: What does the war in Ukraine reveal about the future of warfare and how should we adapt to it?
Ylli Bajraktari: Ukraine offers profound insights, validating many 3OS assumptions while highlighting industrial-age realities. First, there is a more “transparent battlefield”: Ubiquitous sensors (drones, commercial satellites) make surprise difficult and massed formations highly vulnerable. Survival requires dispersion and deception. Second, there is the “drone revolution”: low-cost drones (FPV, commercial UAVs) have democratized access to airpower and ISR. Third, the primacy of fires and EW: artillery and precision strike dominate, but control of the electromagnetic spectrum is essential for survival and targeting. Fourth, mass still matters: warfare consumes vast amounts of munitions, emphasizing the need for a robust defense industrial base. Finally, there are adaptation cycles: the ability to adapt technology and tactics rapidly (measured in weeks, not years) is crucial.
How we should adapt? First, I think we need to accelerate “Replicator”: Ukraine validates the need for large numbers of low-cost, attritable systems. We also need to invest in resilient C2 and EW: develop networks that can operate in a contested EMS. Additionally, there is a need to strengthen the industrial base: ensure the capacity to produce munitions and equipment at scale. Finally, we need to integrate commercial tech: develop faster pathways to integrate commercial technology into military operations.
Octavian Manea: Looking from today’s perspective, what should an Offset X type of strategy be about in order to restore the credibility of deterrence?
Ylli Bajraktari: “Offset X” should not be about discovering the next technological revolution. The technologies identified in the 3OS (AI, autonomy, human-machine teaming) are still the defining technologies of this era. Instead, Offset X must be about operationalizing and scaling the 3OS at speed. The credibility of deterrence rests on demonstrated capability, not just concepts.
The key elements of Offset X should include, first, achieving decision superiority (making JADC2 real): the highest priority must be delivering a functional JADC2 system. We must develop the tools and concepts to sense, understand, decide, and act faster than the adversary, even when communications are degraded.
Second, mastering human-machine integration at scale: moving beyond individual systems to integrating AI and autonomy across the entire force, optimizing battle management and decision support.
Third, revolutionizing the defense ecosystem (speed and scale): the current acquisition system is too slow. Offset X requires a new model emphasizing speed, agility and the seamless integration of commercial technology. We need to move from “inventing” to “fielding” rapidly.
Fourth, dominating the software-defined battlefield: future conflicts will be won by those who adapt their algorithms fastest. Prioritizing software development and the ability to rapidly update systems in the field is crucial.
Finally, redefining resilience: moving beyond platform survivability to resilience in the network, logistics (contested logistics) and the industrial base.
