The Week of July 20, 2026: The economics of drones, social media, tariffs, and much more.
The Week of July 20, 2026: The economics of drones, social media, tariffs, and much more.
The nature of warfare is being changed by drones, and AI is not far behind (if it is behind). In a recent article for the New York Times, W.J. Hennigan discusses both:
The two types of next-generation drones being tested over Southern California… built for the Air Force by Anduril and General Atomics Aeronautical Systems, can fly entirely on their own. They make hundreds of split-second decisions to maneuver, evade and attack enemy targets, based upon countless hours of flight data and computer processing.
Human involvement is limited to determining the mission, starting or stopping it, and deciding whether, so to speak, to press the button.
Hennigan writes that drones of this type flew for the first time last year. Anduril also fired its first AI-assisted (to use an imprecise term) missile last week, and “General Atomics says it is not far behind.”
The Defense Department has “requested $54.6 billion in next year’s budget for the Defense Autonomous Warfare Group, a body set up to coordinate all autonomous weapons programs across U.S. forces.” That is up from $226 million in fiscal 2026, quite a large leap, although $1 billion is the base budget. The rest would come out of the more flexible “reconciliation pot,” a lump sum totaling $152 billion provided for in the One Big Beautiful Bill.
The United States is hardly alone in developing AI-operated weaponry. As Hennigan notes, China is pouring billions into this area, and as far back as 2017, Vladimir Putin declared that whoever leads in A.I. “will become the ruler of the world.” Meanwhile:
Prototypes of smaller, rudimentary A.I.-assisted weapons have already been reported in action in Ukraine, Libya, Azerbaijan and Gaza. In a few years, we could see swarms of self-guiding robots not only in the skies, but also at sea and on the ground.
Hennigan laments the failure of “the international community” to establish limits on AI- weapons. But could this fantasy family ever have done so? Previous international weapons treaties have relied, in part, on the detectability of the weapons deployed. A nuclear explosion is hard to miss, the use of AI rather less so. Moreover, the United States should not commit to a treaty that it will respect, but its rivals will not.
Hennigan argues that “there are practical paths to create international red lines that will make [AI-equipped weaponry] safer — for instance, barring such systems from independently selecting and firing on human targets.” But to trust that Putin, Xi or their like would heed any such boundaries is asking for catastrophe.
It would be more realistic to establish more advanced “hotlines” designed to avoid an accidental Armageddon. And although the United States could never ensure that other powers do likewise, it will almost always be in our self-interest to require a human finger on the button. Even where that might not be appropriate (in a dogfight between AI-directed unmanned drones, say), there should be an off-switch.
Automated drones will affect the equations of war.
Hennigan:
Fear of body bags has been one of the greatest deterrents to conflict in history. Mass-produced, expendable robots do not carry the human or financial costs that have tempered the urge to go to war for centuries.
This is so, but drones, whether “old school” or semi-autonomous, can, as in Ukraine, act as an effective force-multiplier for smaller forces against a larger aggressor.
As I wrote in an article last month:
Ukraine’s drone program, a remarkable demonstration of what ingenuity, improvisation, flexibility, and the private sector can achieve, has played a key part in slowing the Russian advance in the east. I asked one veteran how fighting had changed since 2022. After the drones started appearing in force, the battlefield had, he said, turned into “hell.” “There was nowhere to hide.” The Ukrainians may have the edge, but the Russians have their drones too.
The result has been to create a “kill zone” over and beyond the line of contact where the Russian and Ukrainian armies meet (and any no-man’s-land in between), making any movement in this area extremely dangerous and creating something very close to stalemate.
Ukraine now claims that, mainly owing to drone attacks, Russian troop losses (killed or seriously wounded) are running at about the rate it is thought Russia is able to recruit, around 30,000 a month. As technology improves, so the kill zone widens, from ten miles or so to, in some places, as much as 30 miles. This is taking Ukraine closer to establishing what its former defense minister Andriy Zagorodnyuk has described as the “strategic neutralization” of Russia on the ground. This could mean that instead of being held at bay by conventional and bloody attritional fighting, Russia’s ability to take the offensive on land is, as it already is at sea, frozen.
At sea? Russia had an important naval base at Sevastopol (held under a long lease) in the Crimea. Once Moscow occupied the peninsula in 2014, it appeared set (a wary eye on Turkey aside) to control the Black Sea, but that is not how things have worked out. Writing in the Washington Post, Brady Africk:
Much of the [Russian] fleet has been driven from the Crimean Peninsula, and Russia must now dedicate resources to coastal helicopter patrols to hunt for Ukrainian drones. These helicopters are hunted themselves, threatened by naval drones armed with air-defense missiles. Ukraine is also pioneering unmanned systems that travel completely submerged, a technology the United States, China and other countries are developing.
Drones are not only in the air or underwater. Ukraine’s unmanned ground vehicles are trundling onto the battlefield too. In a robot Thermopylae this winter, a lone armed UGV held off Russian attacks for six weeks. Ukraine has been on track to produce 25,000 UGVs in the first half of 2026. Their UGWs’ range is increasing and so are the uses to which they are put.
Strategic neutralization is a strategy with some potential in countries large enough to host it. With adequate preparation, it ought to be a real deterrent.
It is unsurprising that a drone race is underway, sped even faster along by considerations that blend money and munitions. Thus, the U.S. Air Force is under-equipped to match China’s rapid buildup. According to Hennigan, it views a fleet of robotic drones as a way of keeping pace at a fraction of the cost:
The nearly $1 billion that the Air Force wants for the C.C.A. program next fiscal year is a pittance compared with the cost of high-end fighter jets, which war planners see as continuing to play a prominent role in retaining military dominance. The Air Force wants to integrate at least 1,000 autonomous drones, at an estimated cost of around $25 million to $30 million apiece, into its operations, with the first handful produced next year.
Writing for NRO, Luther Abel crunches some more numbers. The heart of a carrier strike group is, as its name suggests, an aircraft carrier. One function of the ships (and jets) that accompany the carrier is its defense. Smaller craft, such as that on which Luther was serving, would be expected to sacrifice themselves so that the “floating nuclear airfield” (the carrier) survived.
As Luther (writing in this case about Anduril) describes it, what the company “aims to do is swap out the 200 souls aboard a frigate and replace them with swarms of sacrificial drones/chaff/missiles launched from autonomous motherships,” larger (unmanned) rotorcraft. Saving those sailors’ lives is good, but Luther does some cold-eyed “napkin math”:
Each sailor’s family would be due a minimum of $100,000 in the event of the ship’s loss. So $20 million and the public’s outrage at the very least but, more likely, something north of $250 million after all survivor benefits are factored. And that’s just for the loss of our most basic ship class with the smallest crew.
One interesting detail mentioned by Luther (who will shortly have an article on drones in NR) is that Anduril is manufacturing equipment on, essentially, spec. “They use VC funds to make products they think the military needs and only then sell them, which is an inversion of the procurement status quo since the 1990s.” The New York Times’ Hennigan commented on the same thing, noting that Anduril had already constructed a one-million-square-foot production factory to mass-produce the YFQ-44A (a jet-powered semi-autonomous combat drone designed as a “wingman” for piloted fighters) “with just a handful of confirmed U.S. military orders.” Clearly the company is confident that it can find a market for its products. That looks like a decent bet.
In Ukraine, a truly novel turn in military procurement has accelerated drone development. In his blog, Austin Vernon explains one key element, which harnesses market disciplines to pick winners and foster innovation. Brigade-level units buy drones and other equipment directly from manufacturers via the online “Brave1” marketplace, where over 2,500 items are on offer. The “currency” used is converted from “combat points”. These are earned by a unit for video evidence of the destruction of Russian forces or materiel. In 2025, the highest points were awarded for eliminating a mobile rocket system, but scoring may change based on operational needs.
Vernon writes that Ukraine may produce seven million drones in 2026:
They come from dozens of manufacturers and many more unique models that constantly change. The competition and specialization are extreme. Such diversity probably would not be possible without the [Brave1] marketplace.
Necessity is the mother of unorthodoxy. Much of this had been the work of Mykhailo Fedorov both as defense minister and in an earlier ministerial job where he had been charged with “digitalizing” the sclerotic Ukrainian state, a role that had spilled over into the military. It was no straightforward task bureaucratically or financially. As Fedorov explained to The Atlantic’s Simon Shuster, “we had no financing at first . . . We were going around, asking for money from businesspeople, fundraising on Facebook.” The first attack-drone units only started operation in 2023, a year after Russia’s “full-scale” invasion had begun.
Temperamentally and intellectually, Fedorov resembles a Silicon Valley-style disruptor, but he also seems to have concluded that Ukraine had no choice but to change course. It didn’t have the people or the resources to win a war of attrition. Strategically, it would be at a disadvantage if it stuck to the old ways. A small “Soviet” army will lose to a large “Soviet” army. Fedorov’s bottom-up, survival-of-the-fittest procurement regime was the antithesis of a conventional top-down military approach, but, Shuster explains, “Fedorov saw that system as far too cumbersome and prone to corruption. It could not deliver weapons fast enough, he argued, and it did not cater to the ever-changing requirements of soldiers at the front.”
Oleksandr Syrskyi, the commander-in-chief of the Ukrainian military, regarded relying so heavily on automation as too risky, as was upsetting the chain of command. As Shuster sums up Syrskyi’s thinking:
In any war, the top generals need to synchronize the forces under their command to make them function as a single organism, with weapons and tactics that work in concert with one another. The nature of the war in Ukraine has made that challenge particularly daunting . . . When junior officers can decide how to equip their men, keeping everything in sync across the battlefield becomes more difficult. The lower ranks might appreciate the freedom and adaptability of this arrangement, but their commanding officers in Kyiv do not.
Syrskyi is no Soviet dinosaur. In Fedorov’s words, he had “saved” Ukraine in 2022. “He participated in the Kyiv operation, the Kharkiv operation, Kherson. We cannot underestimate such a commander.” And yet Syrskyi appeared unable to find a way to work happily with Fedorov. I was in Kyiv in June, shortly before this conflict led to major trouble. In retrospect, the only hints of it were comments on how tricky things had been before the Fedorov reforms. Rather, discussions, admittedly with people on the drone side, focused on how well things were going now, something which Shuster backs up:
Ukraine now produces several million drones a year, easing its dependence on foreign military aid. Most of the credit for that belongs to the intricate clockwork of drone factories, start-ups, software developers, training academies, and military innovators that help keep the Ukrainian war machine running.
Unfortunately, Syrskyi seemed unwilling to concede that point and just to concentrate instead on ensuring that, in a period of dramatic change, some older but still essential military needs were not neglected.
The tension between Fedorov and Syrskyi reportedly grew too much for Zelensky. He forced out Fedorov, a curious choice (and an inducement to dietrologia), which set off a major political crisis. After street protests, Zelensky replaced Syrskyi with a new Fedorov-friendly commander-in-chief, Mykhailo Drapaty. Yevhenii Khmara, who has a background in the security services as well as the military, has been appointed interim defense minister and may, if in less clear-cut fashion, align with Fedorov’s thinking. What happens next and what (if any) government job Fedorov will hold remains uncertain, but Ukraine would do well to stick with its present course.
Beyond the battlefield, one advantage Ukraine is deriving from its expertise in drone warfare is that it is no longer just a supplicant; it has something to offer the West (beyond — no small matter — acting as its frontline), and not just the drones themselves.
Dan Driscoll, the U.S. Army secretary, recently in Kyiv:
Ukraine’s Delta common operating system, their modular open-system architecture command-and-control system, is absolutely incredible. It fully integrates every single drone, every sensor and every shooting platform into just one single network. Ours does not.
Reuters, July 22:
Ukraine has agreed to export drones to the United States to participate in the Pentagon’s Drone Dominance program, a source familiar with the matter told Reuters on Wednesday.
The agreement comes as Kyiv and Washington have been negotiating a wider drone deal for Ukrainian unmanned aerial vehicles, whose success in the war against Russia has driven high demand from other countries to learn from their experience and purchase models for testing . . .
Ukraine has already signed drone deals with a number of European and Middle Eastern states. Kyiv has also moved to loosen wartime restrictions on its arms exports, although the government still retains a tight hold on what is sent abroad.
Drone Dominance is, Reuters explains, “a billion-dollar competition run by the U.S. Department of Defense” in which manufacturers of small attack drones are invited to put their machines to the test. The first round was won by a drone jointly produced by Skyfall, a Ukrainian company, and Skycutter, a British company. Their partnership is yet another example of the internationalization of Ukraine’s drone industry.
Skyfall is perhaps best known for its Vampyr heavy-load drone, a craft unaffectionately nicknamed Baba Yaga by the Russians after a terrifying creature from their folklore. Ukraine is aware of how its drones are allowing Kyiv to punch above its weight diplomatically as well as militarily. A small collection of its drones can be seen just past the entrance to the foreign ministry in Kyiv, including a Vampyr.
Back to the wider drone race, which continues apace. One aspect of Ukraine’s drone war has been its use of increasingly long-range drones to wreak havoc on Russia’s infrastructure. That raises the question of how to defend against drones. The Ukrainians are themselves naturally using whatever comes to hand, from sophisticated new systems to, amazingly, Maxim guns. But their attacks on Russia are a reminder of drones’ asymmetric power, as are Iranian attacks on a Saudi base that hosts U.S. troops. The technological evolution of anti-drone defenses will be a topic for another time, as will an examination of potential vulnerabilities. As a spoiler, the latter include airfields, and the threat to them is one that China is taking seriously. The authors of a 2025 Hudson Institute report estimated that the amount of concrete used by China to improve the resilience of its air base network could pave a four-lane interstate highway from Washington, DC, to Chicago.
There’s a message there.
The Capital Record: Sound & Vision
We released the latest in our series of podcasts, the Capital Record. Follow the link to see how to subscribe (it’s free!). The Capital Record, which is hosted by financier David L. Bahnsen, makes use of another two formats to deliver Capital Matters’ defense of free markets. The original podcast continues, but if you want to watch David talk, please click on the YouTube link.
Michael’s Gift Stores and the Gift of Private Equity (Podcast/YouTube)
Critics of private equity love to cite obscure or outlier stories of companies that failed and had private equity investment as some sort of proof for the evil of private equity. The arguments are as financially cogent as they are honest, which is to say, they are idiotic and riddled with lies. But in addition to the lies the Elizabeth-Warren class tell about private equity all the time are the clear and obvious counterfactuals to their hideous narrative that they purposely ignore. Today, we are going to not ignore one such case. We are going to tell the story of Michaels, and explain why private equity derangement syndrome has to ignore this case, and the 90 percent-plus of other examples just like it.
The Capital Matters week that was . . .
Banking
President Trump has made affordability a central priority of his second term. Tax relief is putting more money back in the pockets of working Americans. The administration has attacked unnecessary regulations that raise the cost of housing, energy, and everyday goods. The aim is straightforward: reduce barriers, increase competition, and let consumers keep more of what they earn.
That is why recent reports about the Consumer Financial Protection Bureau’s (CFPB’s) forthcoming Section 1033 open-banking rule are so troubling . . .
Drones and Data Centers
Back in July 2025, Christopher S. Chivvis and Sam Winter-Levy of the Carnegie Endowment for International Peace wrote in the Washington Post that a Trump-administration-approved plan to build massive artificial intelligence data centers in the United Arab Emirates and Saudi Arabia was a deeply flawed idea. They warned that the plan would “put a central node of the world’s future AI infrastructure in locations highly vulnerable to attack from hostile actors — and at a time when the risk from cheap drones is intensifying” . . .
Artificial Intelligence
The launch of the Kimi K3 AI model by Moonshot, an early-stage Chinese AI company, is yet another reminder, as if one were needed, that the AI race between the U.S. and China is no foregone conclusion . . .
Housing
Washington’s 21st Century ROAD to Housing Act went into law on July 11 without President Trump’s signature. The benefits of that bill are questionable, and Trump hoped to incentivize Congress to also pass the SAVE America Act, which didn’t happen. This federal housing bill gives governors and state lawmakers an opportunity to think before copying one of Congress’s worst ideas . . .
Drones
Anduril Industries, arguably the Apple of this new generation of venture-capital-funded defense tech firms, just released its latest weapon system: Thunder. An autonomous attack rotorcraft — think an unmanned Falcon from Halo or the chick of a V-22 Osprey — Anduril’s Thunder is being sold as the multi-tool wingman for Apache attack helos and the forthcoming Cheyenne II from Bell (co-creator of the V-22). The added payload, load-out, and protective envelope that such slavecraft (in the engineering sense) could offer is enticing . . .
The States
Paul Schecklman & Paul Wassgren:
America is sprinting into its next era of economic upheaval. AI-driven automation, reindustrialization, supply chain realignment, and historic demographic shifts are occurring rapidly and simultaneously. States that embrace these changes by positioning themselves to meet the moment with pro-growth policies will attract capital, industry, and talent that will define the next generation of prosperity. States that do not will simply be spectators to others’ meteoric rise . . .
Desalination
The operating agreements that govern Colorado River allocations expire at the end of this year, and negotiations between the seven states that share the water are stalled. With the federal government poised to step in and impose an agreement it formulates, the states — Colorado, New Mexico, Utah, and Wyoming in the upper basin, and Arizona, California, and Nevada in the lower basin — have asked for more time . . .
Social Media
In an effort to address a perceived youth mental health crisis, online kids’ safety regulation is on the march in Congress and around the globe. But these government interventions bring with them serious privacy and security trade-offs for internet users of all ages. They are based on unproven hypotheses and fundamentally misunderstand the locus for solving the problem. Whatever the cause of current teen angst, parents, not governments, are the answer . . .
Tariffs
After the Supreme Court struck down tariffs introduced under the International Emergency Economic Powers Act, the president responded by announcing a new global tariff of 15 percent. This was based on Section 122 of the Trade Act of 1974, which allows for the application of various Band-Aids to problems arising out of large and serious U.S. balance-of-payments deficits. The administration’s use of Section 122 to impose tariffs (which was unprecedented) has come under legal attack. But at this point that’s moot (apart from the tricky issue of the tariffs that have already been paid), as Section 122 tariffs could last no longer than 150 days, a period that has just expired . . .
To sign up for The Capital Letter, please follow this link.