The Pentagon Must Undertake Its Own Transformation

By Bing West

Tuesday, July 7, 2026

 

Note: The following excerpt is adapted from Bing West’s book Cat 5: The 2033 War.

 

America is on the path to spend the smallest share of GDP on defense since 1937, a year when the country was trapped in the Depression and wholly unprepared for the coming world war. Over the next five years, the White House and Congress propose spending annually 2.7 percent of GDP on defense. Its real purchasing power of the military may drop to 1.7 percent of GDP, as debt service further squeezes the Pentagon budget.

 

Our forces are shrinking, while China’s are expanding. The Pentagon cannot continue with current business practices and weapons selection. Avoiding defeat in the next war requires a radical break with past procedures. The Pentagon must take two decisive steps: 1) slash its vast army of contractors and 2) redirect those funds into inexpensive, AI-enabled drones and other unmanned systems.

 

1. Fire the support contractors.

 

About 650,000 civilian contractors provide services to our military, at a cost of $250 billion, a quarter of the entire Defense budget. These are not the workers building ships and aircraft. Instead, service contractors perform everyday tasks, such as maintaining computers, delivering supplies, patching communications, training the troops, and so forth. Since 1980, America’s combat arms forces have shrunk by 37 percent to about 400,000, while the contractor workforce supporting them has increased by 70 percent. The surge of service contractors during the Iraq and Afghanistan wars persisted long after those wars ended, becoming the new normal.

 

Each service contractor costs taxpayers roughly $450,000 per year. That astonishing figure includes the overhead of enormous corporations with layers of executives, supervisors, and administrative staff, all adding profit margins on top of salaries. Believing that such a structure makes military sense requires the suspension of disbelief. As Lewis Carroll wrote, one can believe “six impossible things before breakfast,” but believing in the efficiency of the Pentagon’s contractor system remains impossible both before and after breakfast.

 

Comfort always expands. In Vietnam, feeding a rifleman with C-rations cost $8 a day in constant dollars; in Afghanistan, feeding the same rifleman MREs cost $60 a day. In Vietnam, contractors were 10 percent of troop strength. In Iraq and Afghanistan, contractors were equal to or greater than the number of U.S. forces. This carried over after the wars into the standard way of doing business, caused in large measure by the fits and starts of Congress not passing legislation in a predictable way. These service contractors are employed not as individuals, but under massive contracts labeled “communications” or “logistics.”

 

A contractor doing the same work as a GS-15 civilian or an O-6 colonel often costs twice as much. Of the $250 billion in mundane contractor support, at least $40 billion must be cut, with the savings applied to the procurement of drones and other unmanned systems.

 

2. Buy a million AI unmanned systems.

 

AI-enabled unmanned systems have changed the face of 21st century war. Conventional warfare has been transformed by the commoditization of digital technologies. In the Ukraine war, low-cost, off-the-shelf unmanned weapons changed the land and sea battlefields. Russian ships have ceased sailing in most of the Black Sea. Drones have struck Russian airbases thousands of miles apart, and account for 70 to 80 percent of the frontline casualties.

 

Ukraine has made drones central to its war effort, devoting roughly 30 percent of its defense budget, while the U.S. military treated them as niche enablers. In Fiscal Year 2025, the Department of Defense budgeted $25.2 billion on drones and autonomous systems — about 3 percent of its total budget. Allocating 3 percent to the most disruptive category of modern weapons is a denial of battlefield reality. We lag far behind.

 

In 2025, the U.S. Army and Marine Corps had exactly zero drones deployed in their 9,000 combat arms squads. In the commercial world, drones are a commodity, costing $500; U.S. military drones cost $50,000 per unit. Drones at the squad level should cost less than $500. They are munitions to be fired like mortar shells, nothing more and nothing less. Our exorbitant costs have placed us at a steep disadvantage.

 

Ukraine uses a tiered drone ecosystem that ranges from very low-cost disposable systems priced between $200 and $1,000, to mid-range strike drones and unmanned surface vessels costing $12,000 to $30,000, and to long-range loitering munitions priced at around $50,000. This ecosystem reflects a strategy based on speed, volume, simplicity, and adaptation, rather than the U.S. preference for exquisite, integrated, and overpriced systems.

 

The emphasis must be on driving the costs of drones down. This requires iterative competitions among many small firms and government–private sector partnerships, tied to real-time feedback from the warfighters. Secretary of Defense Pete Hegseth has authorized the operating forces, such as brigade commanders, to acquire cheap unmanned systems without having to seek permission from higher headquarters. That is the future.

 

Going directly to the commercial market reduces overall costs because dozens of startups will offer their wares at competitive prices. Several hundred brands of shoes are sold annually in the U.S. We don’t demand that everyone wear the same black shoes. Similarly, there can be dozens of suppliers of unmanned systems. Using this diversified approach, Ukraine produced 4.5 million drones in 2025. The concurrent design and construction of different drones will result in failures. That is offset by funding multiple teams, all inserting rapid software updates. For U.S. forces to deploy a million AI-enabled drones and unmanned naval vessels within two years is feasible in technical terms. No country can match the ingenuity of our software engineers.

 

To shift to multiple inexpensive systems threatens not only the Defense Department’s entrenched contractor oligopoly but also the political interests of Congress. The major defense corporations donate millions to favored elected representatives, and those representatives in turn protect the corporations by injecting thousands of restrictive provisions into procurement laws. “We have 1,500 line items on things we need to buy,” a frustrated Secretary of the Army Daniel Driscoll said in 2025. “We are in a holy war [against Congress] over 1 percent of our budget, to have the flexibility to buy different makes.” Tanks, surface warships, and manned aircraft consume large portions of the budget, each protected by a political constituency that views any shift toward unmanned systems as a threat. Taking funding away from a legacy platform antagonizes presidents, angers congressional committees, and assures opposition from districts tied to old production lines.

 

The military services are also restricted by their own institutional conservatism. Every service jealously defends its legacy systems. Because we are not at war, unmanned systems cannot prove that they offer more warfighting value than the platforms they threaten to replace.

 

In 1938, the best minds in France, England, Russia, and Germany went in different directions in procurement. France defended by engineering, Britain began building fighter aircraft, Russia mass-produced inferior weapons, and Germany manufactured mechanized divisions. Each nation pursued its own theory of victory. Today, our four services reflect that same divergence. Each moves along its own procurement path with inadequate regard for a unified strategy against China, our primary adversary. The Army (20 percent of the Pentagon’s budget) struggles to define its role in the Pacific. The Marines (5 percent of the budget) have no coherent strategy and remain small in absolute terms. The Air Force (31 percent of the budget) is investing in long-range missiles launched from aircraft to strike deep inside China, targeting ports, air bases, command centers, and missile installations.

 

The Navy (26 percent of the budget) is focused on a battle against China at sea. White House and congressional support is strong for large, expensive warships and the thousands of shipbuilding jobs they create. President Trump has advocated creating “pocket battle ships” as the expensive centerpiece of what he calls a “Golden Fleet.” This reflects an emotional reverence for World War II, when armadas of manned ships sailed to victory. In 2025, the Navy allocated $20 billion to surface warships and only $1 billion to unmanned vessels and drones. That allocation defines a mindset out of step with modern warfare.

 

Ships funded in 2025 will not enter service until 2035 and will deploy for the rest of this century. It is difficult to imagine that massive carriers — each larger than three football fields — remaining survivable in the battlespace of 2040, let alone 2100. Retired Admiral James Stavridis, former Supreme Allied Commander in NATO, has warned that maritime warfare has reached an inflection point. “We are at an absolute pivot point in maritime warfare,” he wrote. “Big surface ships are highly at risk to air, surface, and subsurface drones. The sooner great-power navies like that of the United States understand that, the more likely they are to survive in major combat in this turbulent 21st century.”

 

China’s surveillance and missile networks have turned the Western Pacific into a transparent battlespace. Any carrier, amphibious ship, or large surface combatant nearing Taiwan will be detected, tracked, and targeted by thousands of precision weapons. The mathematics of modern naval warfare is unforgiving. Defensive missiles cost millions of dollars each; the drones and missiles attacking them cost tens of thousands. The United States must spend $1 million to kill a $50,000 attacker. This is a losing exchange, made worse by the finite number of defensive missiles each ship can carry.

 

China will expand its missile arsenal by 100 percent over the next decade, while increasing its fleet by 17 percent and its airpower by 25 percent. By elementary quantitative measures, the U.S. Navy will confront a stronger China in 2033.

 

This trend can be reversed. Unmanned surface ships between 200 and 300 feet long can carry hundreds of drones or missiles and can be built for a fraction of the price of a Ford-class carrier, which costs between $13 billion and $15 billion. Taking into account both capital expenses and attrition rates per strike, unmanned vessels with AI-guided drones and missiles generate destructive power at roughly one-third the cost of carrier-based aviation. The United States cannot afford to treat the 21st century like an extension of 1945. The mission of both carriers and unmanned vessels is to deliver ordnance on target; the cost per strike heavily favors unmanned platforms.

 

Absorbability — meaning the capacity to take losses and continue fighting — is the reality of battle. China is willing to absorb heavy losses of human life; we are not. Unmanned systems compensate for the vulnerability of manned platforms by absorbing losses at little human cost, while carrying vast numbers of munitions and sustaining strikes over time. If an attacking force can launch thousands of cheap drones and missiles each day, it can overwhelm an adversary whose interceptors cost ten times more. The defender’s magazines empty long before the attacker’s do.

 

The historical precedent for this shift is clear. At Agincourt in 1415, the wealthy French knights brought tradition based on prior victories; the lowly English longbowmen brought cheap mass firepower. English archers remained plentiful, long after France had run out of knights. In 2033, carrier battle groups are the knights — majestic, expensive, impossible to replace — while surface and aerial drones are the archers — disposable, numerous, and replaceable as a commodity. The historical comparison to Agincourt is obvious: The side that can absorb losses at 1/20th the cost wins the battle.

 

In summary, the Pentagon must undertake its own transformation. Congress will not significantly increase the defense budget. Fortunately, the United States military has a tradition of self-criticism, vigorous internal debate, and adaptation. The Pentagon must cut back the bloated contractor corps, scale down legacy platforms, and redirect 20 percent of procurement toward unmanned AI systems. To their credit, the senior civilian Defense officials understand this. Pete Hegseth and Deputy Defense Secretary David Feinberg have fundamentally altered the procurement system, sweeping out scores of stultifying regulations, opening up competition, and insisting upon rapid innovation in conjunction with best civilian high tech practices.

 

Radical change in procurement practices has begun. But by 2033, interest on the national debt is likely to consume one-quarter of all federal revenues, while entitlement spending absorbs nearly all of the remainder. A profligate Congress must borrow roughly $2 trillion annually to finance national defense and essential public infrastructure. As debt mounts, investors will demand a risk premium, diverting capital from productive private investment, slowing economic growth. It will be a remarkable achievement if sufficient autonomous warfare systems are produced to maintain our military dominance while the Defense procurement is squeezed and legacy systems are protected.