WASHINGTON >> When a North Korean missile test went awry Sunday, blowing up seconds after liftoff, there were immediate suspicions that a U.S. program to sabotage the test flights had struck again. The odds seem highly likely: Eighty-eight percent of the launches of the North’s most threatening missiles have self-destructed since the covert American program was accelerated three years ago.

But even inside the U.S. Cyber Command and the National Security Agency, where the operation is centered, it is nearly impossible to tell if any individual launch is the victim of a new, innovative approach to foil North Korean missiles with cyber and electronic strikes.

Bad welding, bad parts, bad engineering and bad luck can all play a role in such failures — as it did in the United States’ own missile program, particularly in its early days. And it would require a near impossible degree of forensic investigation to figure out an exact cause, given that the failed North Korean missiles tend to explode, disintegrate in midair and plunge in fragments into faraway seas.

But this much is clear, experts say: The existence of the U.S. program, and whatever it has contributed to North Korea’s remarkable string of troubles, appears to have shaken Pyongyang and led to an internal spyhunt as well as innovative ways to defeat a wide array of enemy cyberstrikes.

By all accounts, the program that President Barack Obama stepped up in 2014 has been adopted with enthusiasm by the Trump administration. President Donald Trump’s national security aides are eagerly hoping that the Chinese, among others, will get North Korea to freeze or reverse its program. Yet they have no compunctions about using this new class of weapon against missile tests that the United Nations has already prohibited.

Speaking in Moscow last week, Rex W. Tillerson, the U.S. secretary of state, was put on the defensive by a Russian reporter who challenged U.S. complaints of interference in the American election, despite Washington’s cyber attacks against Iran’s nuclear program and North Korea’s missile program. “Cybertools to disrupt weapons programs — that’s another use of the tools, and I make a distinction between those two,” Tillerson argued back, without specifically confirming their use against Pyongyang.

Perhaps taken by surprise at the question, Tillerson never took the next step to voice the argument that some of his Trump administration colleagues make in private: that since the U.N. Security Council has banned North Korean missile tests, any effort to interfere with them would have some basis in international law.

“When you look at what is emanating out of North Korea,” Michael Chertoff, a former secretary of homeland security who now runs a cyberconsulting group in Washington, said Tuesday at the Carnegie Endowment for International Peace, “I have sympathy for the argument that anything we can do to stop an unpredictable person from using nuclear weapons is worth trying.”

But the question for the United States’ intelligence agencies is whether this new tool is as effective as many have hoped. While billions of dollars have been poured into new offensive cyberweapons, touting a success in thwarting North Korea — whether it is real or imagined — can be turned into an argument for more.

It is a particularly difficult question in light of Sunday’s botched test, because it is still unclear exactly what missile was launched. By nature, missiles teeter on the brink of failure, and new designs are often accident prone. At their best, missiles are dense welters of pipes, engines, valves, pumps, volatile fuels, relays, explosive bolts, wires, sensors and circuit boards that suddenly emit blistering flames and roar skyward with such shattering violence that they often quickly hit the breaking point. Things can easily go wrong, and frequently do.

But even by those measures, the North Koreans are having a rough time, and it has gotten a lot rougher since the United States accelerated its sabotage program.

In the annals of rocketry, experts say, roughly 5 to 10 percent of developmental test flights go awry. That holds even for such high practitioners of the art as the billionaires Jeff Bezos, the founder of Amazon, and Elon Musk, the founder of Tesla, who are now racing to redefine the future of spaceflight. (By contrast, U.S. commercial airline flights have a success rate of more than 99.9999 percent. And when crashes do occur, it can take investigators weeks, months or even years to identify the cause.)

But the sheer frequency of North Korean missile mishaps suggests that sabotage lies behind at least some of the recent failures.

So does the timing. Typically, countries encounter high failure rates when they start their rocket programs. As the programs mature, and engineers gain experience, spectacular failures decline and success tends to become a habit. In North Korea, the situation has been the exact reverse.

By and large, the North was a reliable maker of missiles in the 1980s, ’90s and into the 2000s. The government sold its missiles to Pakistan and Iran, among others.

Then came the effort to launch the Musudan, an intermediate-range missile that Pyongyang first displayed in a military parade in late 2010. It was 5 feet wide and 40 feet long — remarkably small compared with the North’s big rockets. But it represented an enormous threat. Carried on a truck, it could be hauled on country roads through forested regions or kept in tunnels, making it easy to hide and, as a target, difficult to find and destroy.

To date, the proven reach of the Musudan makes it the most threatening potential weapon in the North’s emerging arsenal of missiles that might loft nuclear warheads. It is seen as able to hit targets up to 2,200 miles away — far enough to strike the sprawling U.S. base at Guam.

Last year, the North conducted eight flight tests. Only one succeeded, giving the missile an overall failure rate of 88 percent. It was after the last failure that the North Korean leader, Kim Jong Un, was reported to have ordered an investigation into whether the United States was sabotaging his country’s test flights, searching for spies in his system.

Experts say that the best way to slow a program is to send a country scrambling for the causes of failures. “Disrupting their tests,” William J. Perry, defense secretary in the Clinton administration, said at a meeting this year in Washington, would be “a pretty effective way of stopping their ICBM program.”

But more recently, the effectiveness of the United States’ sabotage has grown increasingly uncertain. Some new North Korean missile designs, using solid fuels, have had a higher success rate. Moreover, the North Koreans, as sophisticated cyberoperators, have grown better at defense.

John Schilling, a technical expert on North Korea’s missile program, expressed skepticism Tuesday about the efficacy of the foreign cyberattacks against Pyongyang’s missiles.

“We haven’t seen anything yet pointing to cyber specifically,” Schilling said on a conference call organized by 38 North, a think tank specializing in North Korea at the Johns Hopkins School of Advanced International Studies.

An easier target of sabotage, Schilling added, would be the parts and supplies that North Korea imports to feed its factories that make the missiles.

© 2017 The New York Times Company