Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
Maintaining deterrence in the Indo-Pacific and around the world requires the U.S. Navy to change what it builds and how it fights. Sen. Roger Wicker observed in 2024 that the United States’ approach to fleet design and ship construction is “too small and too old.” The current model of naval power cannot scale at the speed modern war demands. The war with Iran is already exposing the limits. High-end ships are being consumed in sustained operations, munitions inventories are thinning, and replacement timelines for exquisite weapons stretch into years. Against a more capable adversary, such as China, those constraints would not be manageable. China’s advances in its ability to detect, track, and strike ships mean that America’s advanced surface ships — even aircraft carriers — are now profoundly vulnerable.
The Navy’s struggles are putting deterrence and warfighting capabilities at risk. Naval power is as central to U.S. economic and military power today as it was when Naval officer and historian Alfred Mahan published The Influence of Sea Power Upon History in 1890 and coined the phrase that “whoever rules the waves rules the world.”
To build a larger and more capable fleet by 2030, the Navy should double down on investing in, fielding, and sustaining smaller attritable autonomous systems to support Indo-Pacific Command’s Hellscape concept, which will increase short-term deterrence and warfighting capabilities. Further, as the Iran war shows, the American military is short on munitions, and replacing exquisite, expensive, hard-to-produce weapons like Tomahawks will take too long. To complement the existing arsenal, the Navy should aggressively procure lower-cost, long-range munitions. For example, the Navy should follow the example of the Air Force by rapidly moving a low-cost cruise missile into production.
Third, and most disruptively, the Navy should maximize funding for the development and production of what it calls “medium unmanned surface vessels,” autonomous boats between 45 and 190 feet with displacement up to 500 tons. For the sake of using plain language, I will simply call these kinds of vessels, medium autonomous warships, given ever changing terminology and debates about is and is not a ‘drone.’
The Navy should look in particular at lessons learned from the USX-1 Defiant, which was christened on Aug. 11, 2025. Medium autonomous warships should be purpose-built to lower costs and increase efficiency, constructed in yacht yards, smaller shipyards where large speedboats or yachts are constructed, instead of larger shipyards, and made capable of installing vertical launch tubes for missiles or surveillance equipment. Medium autonomous warships should become the Navy’s Collaborative Combat Surface Vessels, the surface fleet version of Collaborative Combat Aircraft in the air. A larger fleet with autonomous ships will make the Navy more resilient and less vulnerable to losing any one particular ship, assist with operations in littoral areas such as the Strait of Hormuz, and also complicate China’s planning by generating more aim points and attack vectors. It will be organizationally difficult for the Navy to shift from focusing exclusively on small numbers of capital-intensive ships to a high-low mix that includes scaling multiple types of autonomous ships, requiring overcoming deeply embedded bureaucratic incentives and pathologies. The upsides for the nation are enormous, and the scale and timeframe of the China threat and pace of technological change means now is the time for boldness.
Threats to the Navy’s surface fleet are growing from multiple domains — particularly from China, but also from Iran, the Houthis, and the many other actors capable of saturation attacks in the era of precise mass. Precise mass reflects the intersection of commercial manufacturing, advances in artificial intelligence and autonomy, and the proliferation of precision guidance technology that now means all actors can generate accurate sensing, strike, and potentially even platforms at lower costs. It’s critical for affordable mass, which Purple Rhombus CEO Mike Benitez defines as “the ability to replace losses as fast as they are likely to occur.” Ground-launched and air-launched missiles — both China’s sophisticated variants and more inexpensive ones deployed by Iran or the Houthis — can increasingly hold surface vessels at risk. China’s Navy has fielded at least six different modern anti-ship missiles over the last 15 years. They include air-launched and sea-launched subsonic and supersonic cruise missiles (YJ-12, YJ-18, YJ-83), ground-launched anti-ship ballistic missiles (DF-21D and DF-26D), and ground-launched hypersonic cruise missiles (DF-17). China is also investing in low-cost, precise mass capabilities more like Iranian and Houthi capabilities that could stretch and overwhelm U.S. defenses, as Iran’s current attacks throughout the Middle East illustrate. The mix of these threats, combined with China’s increasingly extensive sensor network, places Navy assets at increasing risk, especially larger ships.
While undersea superiority is the Navy’s ace in the hole, providing the backbone of U.S. conventional military superiority, China’s submarine forces and anti-submarine warfare assets are also improving. Russia’s sophisticated submarine forces, including the Severodvinsk-class boats, are also a threat. Even if U.S. submarines outmatch their competitors, adversary submarines also present a potent threat to Navy surface ships.
As of May 4, 2026, the U.S. Navy has 291 ships. That total is well below the 355 mandated by the 2017 National Defense Authorization Act and even further below the 381 ships the Navy told Congress it needed in 2024. The Congressional Budget Office estimates that, despite the billions invested by the Biden and Trump administrations into shipbuilding, fleet size will drop further to 283 in 2027, which is also the year General Secretary Xi Jinping charged China’s military with being able to mount an amphibious invasion of Taiwan. Moreover, the Navy is expending massive resources in the fight against Iran, using several years of Tomahawk inventory, and with a third aircraft carrier group now headed towards the Middle East. Increasing the size of the fleet by preventing ship retirements en masse or bringing ships out of mothballs would lead to an older, less capable fleet. That will undermine, not improve, deterrence.
Efforts to produce more ships through traditional shipyards and to repair the ones already deployed are struggling. Nearly every single U.S. Navy shipbuilding project is behind schedule by years and over budget. For example, Virginia-class Block IV submarines and Constellation-class frigates are both 36 months delayed, and the Virginia-class program is currently $17 billion dollars over budget.
Fixing the defense industrial base is essential, but the billions invested by the Trump and Biden administrations to recover from the devastating impact of the 1990s peace dividend will take years to pay off. The United States stopped authorizing the procurement of attack submarines in 1991 with the order of the third Seawolf boat, gapping the submarine industrial base until procurement of the Virginia-class program was launched in 1998. In between, the industry collapsed, with companies merging, workers leaving and retiring, and facilities falling into disrepair. Industry is currently producing 1.2 Virginia-class submarines a year, compared to a requirement to produce 2 Virginia-class submarines a year plus 1 Columbia-class submarine a year.
The Office of Shipbuilding and Maritime Security is another step forward towards revitalizing U.S. shipbuilding, but it too will take time to pay dividends for the Navy. There is a shortage of large shipyards, even if they operate efficiently. There are four outdated publicly owned shipyards. The newest dry dock at any of them, at Puget Sound, was completed in 1962. There are five private shipyards currently building large Navy vessels, and another 15 mostly focused on commercial construction. Nearly all are fully booked at their present capacity level for several years out. Increasing the construction rate, therefore, requires building ships faster, increasing building capacity, or both. Moreover, maintenance on Navy vessels is notoriously slow and expensive, as the decade-long USS Boise repair saga illustrates.
But those are not the only places to build ships. There are 60 additional inactive facilities that once produced Navy ships, submarines, or oceangoing cargo vessels, 86 active smaller shipyards, and eight active yards capable of producing megayachts in the United States. This means there’s an opportunity to dramatically expand autonomous ship construction through the creation of yacht-sized boats under 200 feet or larger boats built in megayacht yards.
Navy leaders already understand, in theory, what is necessary to expand the fleet before 2030, given the length of time it will take to fully address the naval industrial base: invest in robotics, AI, and autonomy. As Adm. Sam Paparo of Indo-Pacific Command writes: “Building capital ships takes years,” arguing for investing in autonomous ships “to increase warfighting capability in the near term … augment[ing] the multi-mission conventional force to increase lethality, sensing, and survivability.”
A new generation of non-traditional contractors is ready to deliver, using additive and advanced manufacturing techniques and more modular designs. Leaps ahead in AI and autonomy for battle management and air power also suggest the time is ripe to construct a larger Navy more capable of sea control and sea denial around the world, especially against China.
The top priority should be scaling a medium autonomous warship as a new ship for the Navy. The Navy’s new call to industry for off-the-shelf solutions is sensible, and the Navy should also look to Defense Advanced Research Projects Agency’s (better known as DARPA) No Manning Required Ship program. When the Navy pivoted away from the behind-schedule, and over-budget optionally crewed Large Unmanned Surface Vessel program, a key driver of its woes and $500 million price tag for the first boat was the requirement for optional manning. Making room for sailors while the ship is underway requires space for the crew, heating and cooling systems, kitchens, beds, and other requirements that significantly increase the cost of the vessel. The DARPA test craft USX-1 Defiant, in contrast, costs approximately $25 million excluding mission systems or weapons.
Large unmanned surface vessels were designed to be 200 to 300 feet with a displacement of 1000 to 2000 tons. The Defiant, which is hitting the water for sea trials, is 180 feet long and weighs 240 metric tons, meaning it can be produced in yacht yards. The ability to produce medium autonomous warships in yacht yards — given how backed up the larger shipyards are with production and the need to focus them on larger vessels and submarines — means there are opportunities for the Navy to scale ships faster at the maximum size that could be constructed in yacht yards.
Yacht yard-sized boats could also carry substantial payloads for surveillance and strike. For example, a boat the size of the Defiant could carry two sets of four-cell launchers launching any missile that fits in a vertical launch system tube. A larger format ship named the Dauntless could carry up to four sets of four-cell launch tubes. Making the ship autonomous, not optionally crewed, generates the space savings that allow for the insertion of launch tubes or containerized surveillance packages.
A promising concept would involve using medium autonomous warships as Collaborative Combat Surface Vessels linked to cruisers or larger ships via the Aegis Combat System, but operating mostly autonomously, with the degree of autonomy increasing as technology improves and inserted via an open architecture approach to the software.
With the Defiant currently undergoing sea trials and a new solicitation out to industry, the Navy should immediately pivot to acquisition using the $2.1 billion appropriated to procurement of medium autonomous warships in the defense reconciliation bill. Medium autonomous warships that can launch weapons or conduct surveillance should then count as surface vessels for the Navy, though this will require changing the Navy’s counting guidance to include larger-format drones. Medium autonomous warships are not just ships on paper to solve a counting issue: They will have growing levels of autonomy and be able to conduct multiple missions, including launching up to 16 missiles, meaning it is appropriate to consider them a ship.
The precise mass opportunities for the Navy do not just come from Collaborative Combat Surface Vessels, but from smaller drones based on the Hellscape concept promoted by Indo-Pacific Command, which involves unmanned surface vessels as munitions rather than ships.
Both shorter- and longer-range surface drones, smaller than medium autonomous warships, are ready to scale now. While not every design will succeed, experimentation and fielding in Central Command today can pave the way for fielding in the Indo-Pacific. Lessons learned from Ukraine in the Black Sea show that smaller naval surface drones can effectively execute one-way attacks and launch attacks of their own, meaning they could increase the probability of success for the Navy in the Taiwan Strait before 2027. Surveillance platforms such as Saildrone and short-range attack surface drones are ready to scale now. For example, AEVEX, producer of the Phoenix Ghost one-way attack aerial drone, produces a naval drone called the Mako. There are recreational craft companies capable of producing small surface drones, whether as single-use for strike purposes or for surveillance.
As the Iran war showed, the Navy — like the other U.S. military services — faces massive shortages in critical munitions, which would undermine its ability to fight at scale for more than a few weeks in the Indo-Pacific. A larger Navy will require low-cost munitions to arm the fleet. The Multi-Mission Affordable Capacity Effector initiative to produce a low-cost cruise missile is promising. Navy and Air Force efforts to procure low-cost cruise missiles received $1.5 billion in the defense reconciliation bill, and $133 million is slated to go to the initiative. The Navy followed this up with a Fiscal Year 2027 budget request for the effort that includes $156 million to procure 353 missiles. The Navy also needs to follow through on the plan to spend $1.6 billion to purchase about 4500 low-cost missiles by 2031, which works out to under $400,000 per missile. Given delays in production and performance issues in past programs, this will require a laser focus to ensure it continues to progress.
At the same time, Central Command’s fielding of emerging capabilities highlights the potential in the Indo-Pacific as well. In the early days of the conflict, Task Force Scorpion Strike fielded the Low-Cost Unmanned Combat Attack System — the American military’s first precise mass system — which can be launched from ships. Reverse-engineered from Iran’s Shahed-136, the Low-Cost Unmanned Combat Attack System has a similar price point and the ability to reverse the cost curve on adversaries. U.S. Central Command commander Adm. Brad Cooper has described the Low-Cost Unmanned Combat Attack System as indispensable for America’s operations. Given how just a month of fighting against Iran has substantially reduced U.S. stockpiles of exquisite, expensive, difficult to produce weapons such as the Tomahawk, naval variants of the Low-Cost Unmanned Combat Attack System, and other one-way attack systems would complement low-cost cruise missiles and ensure the Navy has the weapons it needs for both a shorter-term and longer-term fight in the Indo-Pacific.
Some might argue that increasing the number of autonomous ships and complementing the Navy’s arsenal of exquisite weapons and platforms might undermine deterrence because these ships and weapons will not have the credibility with allies and partners — or adversaries — as ships with people on them. Nonetheless, following the strategy suggested above will increase deterrence for several reasons. First, forward presence is already undermined because the Navy doesn’t have enough ships and cannot do maintenance on existing ships fast enough to keep them ready. Moreover, proposals to increase the size of the existing fleet that rely on existing shipyards will take years, and building a new shipyard would take a decade or more. The question isn’t adding autonomous ships versus a new wave of traditional ships, but whether it’s time for urgent change, given that deterrence is already eroding.
Second, ships that are not useful for fighting China will not enhance deterrence in the Indo-Pacific. The Navy should not sacrifice warfighting effectiveness because having sailors on boats that are at high risk of being sunk as a tripwire somehow makes war less likely. Third, the current demands on the Navy already place too much stress on the fleet, undermining readiness and deterrence. Fourth, deterrence is not the sole responsibility of the Navy: Other services could ramp up other forward presence activities. Finally, a bigger Navy means there will be more American-flagged ships in more places. That alone will increase deterrence.
The war with Iran has also generated real-time operational evidence illustrating the value of precise mass and the threat it can pose to crewed surface vessels and other targets. Iran’s potential use of combined drone and missile salvos and fast attack boats in the Strait of Hormuz has exposed the threat to American and commercial ships in contested littoral environments. The lessons from the Gulf and Red Sea are not merely analogous to the Indo-Pacific problem: They are a direct rehearsal for it. If these systems prove effective against Iranian threats, the case for scaling them for Hellscape and Collaborative Combat Surface Vessels missions in the Taiwan Strait becomes substantially stronger.
To be clear, the Navy’s existing experience with Sea Hunter and the Ghost Fleet Overlord program shows there is work to do to show that Collaborative Combat Surface Vessels will be operationally reliable in adversarial environments that include electronic warfare, GPS denial, and cyber attacks. Investing heavily in technology and operational concepts to address those issues is essential. The risk is worth it for several reasons. Improving autonomy technology means the risk of technology failure is lower than a decade ago, when the Navy decided to pursue an optionally crewed approved for the large unmanned surface vessel. Moreover, the long timelines for improving the production rate for larger warships and submarines means ships capable of construction in yacht yards, like medium autonomous warships, are the only option to improve capacity in the shorter-term. Especially in the wake of the Iran War, the Navy could struggle to meet deterrence requirements in the Indo-Pacific by 2027. The boost to the defense budget also reduces the investment tradeoffs, making now an opportune time to move forward. One might worry about the financial or credibility cost if the program does not succeed, but operating medium autonomous warships initially in lower-threat environments and theaters while progressively expanding their operational envelope is a sensible path to building the track record for trusted confidence in operating them in high-end contested environments. Finally, the well-documented struggles of the current ship construction approach mean the larger risk is not trying something new. The time has come to move towards Collaborative Combat Surface Vessels.
The U.S. Navy is confronting a structural mismatch between the character of modern warfare and the force it has built to fight those wars. The era of precise mass is altering the economics and operational logic of naval power. A naval fleet built exclusively around a small number of exquisite, capital-intensive platforms and weapons is increasingly brittle. Deterrence in the Indo-Pacific will not hinge on marginal improvements to legacy systems over the next half-decade, but on whether the Navy can adapt its force design, acquisition practices, and operational concepts to this new reality. Improving naval deterrence and warfighting before 2030 requires making larger bets and being willing to take risks. The Navy should embrace Collaborative Combat Surface Vessels, small surface drones aligned to the Hellscape concept, and low-cost cruise missiles and one-way attack drones.
Investing in medium autonomous warships, lower-cost long-range munitions, and surface drones is not about affordably generating the volume and diversity of effects necessary to survive and fight in a contested environment. The Navy should treat these capabilities not as adjuncts to the existing fleet but as central components of future maritime power.
The primary barrier to adoption is organizational. The Navy’s entrenched preferences for large platforms and legacy procurement pathways risk slowing the transition at precisely the moment when speed matters most, given the scale of the challenges. China’s military modernization and the proliferation of precise mass capabilities mean that these changes should start today to be able to fight and win both tomorrow and in the future. A more distributed Navy will be a more resilient Navy, but the window for adaptation is narrowing. A Navy that can field a distributed, resilient force incorporating precise mass will strengthen deterrence. The choice is not between tradition and innovation — it is between relevance and obsolescence.
Michael C. Horowitz is the Richard Perry professor and director of Perry World House at the University of Pennsylvania and a senior fellow at the Council on Foreign Relations.
Image: U.S. Navy
