Sean Connery in his role as James Bond.
Credit: X/@Super70sSports
By Samuel Townsend and Amy Zegart
On March 12, America’s spy chiefs gave Congress their annual intelligence threat assessment of dangers facing the nation, but few Americans noticed. A whopping 9,000 people watched the hearings on C-SPAN. Neither the report nor the hearings made the front pages of the biggest US newspapers.
Indeed, each spring for the past 30 years, spy agencies have come out of the shadows to publish an unclassified assessment of the biggest global threats facing the US — giving Americans a rare glimpse into who they are, what they do and how they think. And for 30 years, almost nobody has paid attention. Most Americans think about their intelligence agencies only when they mess up or get embroiled in controversy. It’s like trying to understand football by watching only the fumbles, incomplete passes and penalties.
This selective attention makes informed judgments of intelligence failures both more important and more challenging. They are more important because failures are rare moments when focused public attention can drive meaningful reforms, such as when al-Qaeda’s 9/11 attacks led to the creation of the director of national intelligence and other improvements to counterterrorism efforts.
At the same time, assessing intelligence failures in moments of crisis is more challenging, because that’s when casual observers rush into action — often drawing the wrong conclusions about how intelligence operates, what went wrong, and how to fix it.
Failures in the public analysis of intelligence mistakes stem from four common pitfalls.
The first is the tendency to assume that when bad things happen, intelligence breakdowns must be the cause. Yes, the primary mission of intelligence agencies is providing advance warning of nasty surprises. But preventing those surprises from turning into tragedy takes two to tango: Intelligence agencies have to warn, but it’s up to policymakers to act. Often, the policymaker part of the equation gets forgotten and the term “intelligence failure” gets conflated with the consequences of policy decisions or inaction.
For example, while it is often claimed that the CIA squandered several opportunities to kill Osama bin Laden before the Sept. 11 attacks, the 9/11 Commission Report details multiple instances — including in 1998 at Tarnak Farms in Afghanistan — where high-quality intelligence on bin Laden’s whereabouts and detailed plans to take action were not acted on because of White House concerns over potential civilian casualties.
More recently, when US intelligence agencies took the unusual step of declassifying sensitive intelligence and sharing it with the world to warn that Vladimir Putin was about to invade Ukraine, many world leaders, including Ukrainian President Volodymyr Zelenskiy, didn’t believe it right away. Ukraine has fought with heroic courage over the past two years, but it is worth asking how earlier and more forceful action by Zelenskiy and his European supporters in response to America’s intelligence warning might have made a difference.
A second common pitfall is the search for oversimplified answers. When serious breakdowns in intelligence occur, identifying root causes is essential — and not nearly as simple as it seems. Hot takes on intelligence failures are often wrong takes because they latch onto simplistic explanations based on hindsight. The most common is concluding the danger was painfully obvious but spy agencies were just too bumbling to see it coming.
Intelligence is a business of gray, not black and white. Information on adversary capabilities and intentions is almost never complete, often deceptive, and is almost always consistent with multiple interpretations. Even smoking guns are rarely smoking guns. In the runup to the first Gulf War, for example, imagery clearly showed Iraqi leader Saddam Hussein’s tanks massing near the Kuwaiti border. The question was what he planned to do: to invade or to bluff to gain the upper hand in oil negotiations with the Kuwaitis? It may be that Saddam himself wasn’t sure until he gave the order to invade.
There is an inherent lack of certainty in any enterprise attempting to anticipate the future. Everyone knows it’s difficult to forecast the precise path of a hurricane, pick the winner of the Kentucky Derby, determine the stock price of a new public company, or predict the popularity of a new movie. And in those cases, nobody is battling well-resourced adversaries actively attempting to deceive or conceal information.
What’s more, in intelligence, success and failure aren’t so distinct; intelligence agencies can succeed and fail at the same time. Declassified US intelligence warned the world of Putin’s planned invasion well before his own generals knew — a stunning success. And yet US intelligence agencies also assessed that the Russians would march to victory in days, which turned out to be dead wrong. So sure was the US that Putin would roll through Ukraine, the embassy in Kyiv was evacuated. Intelligence officials conceded afterward that they did not adequately account for variables such as the Russian military’s internal weaknesses or the Ukrainians’ strong will to fight.
A third pitfall is the focus on individual mistakes instead of underlying organizational deficiencies. Especially in moments of tragedy, putting people at the center of a story feels like accountability, and it makes for great politics and news headlines. Yet massive intelligence failures almost never stem from one individual making a single tragic error. Systemic organizational weaknesses and cognitive biases are far more often the cause.
Before 9/11, for example, more than 50 officers inside the CIA had access to information that two suspected al-Qaeda operatives were probably in the US, yet none of them notified the FBI for the next 18 months. The two suspected operatives went on to crash American Airlines Flight 77 into the Pentagon. What went wrong? It wasn’t that all those CIA officers fell down on the job. The real problem was that notifying the FBI of possible terrorists on US soil hadn’t ever before been the CIA’s job. For half a century, the agency had operated with Cold War priorities, procedures and policies.
As a general rule, when that many people make the same bad call, the problem isn’t them, it’s the organization — the structures and incentives that lead people to keep doing the things the same old ways even in the face of new threats. Unfortunately, it often takes a crisis to drive adaptation of the system.
The fourth pitfall stems from the impact of fictional spies. Although US intelligence agencies are engaging more with the public than they used to, spy-themed entertainment is still the primary source of adult education about espionage. Most Americans know deep down that what they see in the movies and on TV isn’t the same as reality, but they don’t know how or how much. From James Bond to Jason Bourne, Carrie Mathison to Ethan Hunt , the entertainment industry loves to portray intelligence officers either as cowboys running wild to stop villainous plots without regard to the law, or as Big Brother who is tracking your every move and conversation with grandma. Neither portrayal is remotely accurate.
The vast majority of intelligence activities are focused on collecting information about foreign adversaries in order to provide timely analysis to policymakers. Of the 18 intelligence agencies that constitute the Intelligence Community (IC), only one, the CIA, conducts covert action — and then only in foreign countries, such as aiding the Afghan mujahadeen against the Soviets, or rescuing American hostages from Iran after the 1979 revolution.
Importantly, covert operations can only be conducted in very specific and highly controlled scenarios . They require written authorization from the president and congressional notification, regular oversight and dedicated funding appropriations.
Similarly, surveillance activities of US intelligence agencies are very different than how they are portrayed in popular culture. To be sure, American intelligence agencies had some dark days involving massive illegal surveillance of Americans in the 1950s to the 1970s. As a result of these abuses, the IC operates under much stricter authorities governing what it can and cannot do — especially with respect to American citizens. Any collection that does happen inside America — whether intentional or incidental — is subject to tightly controlled statutory and judicial oversight mechanisms that have stood for 50 years. To be effective, US intelligence agencies must align with American values, not just American laws. When they stray from those values, Congress can and does take action. When former National Security Agency contractor Edward Snowden revealed information about highly classified NSA programs that stirred national controversy, for example, Congress changed how Americans’ bulk phone records could be collected and instituted other reforms to increase oversight transparency.
Intelligence agencies have been tasked by America’s leaders to collect and keep secrets to protect the nation. They will not always succeed. In moments of failure, learning the right lessons and avoiding misperceptions is essential.
Samuel Townsend is a national security affairs fellow at the Hoover Institution and lead strategist for the Central intelligence Agency's Directorate of Science and Technology. Amy Zegart is a senior fellow at the Hoover Institution, professor of political science (by courtesy) at Stanford University, and the author most recently of “Spies, Lies, and Algorithms: The History and Future of American Intelligence.”
Disclaimer: The views expressed here are those of the authors and do not reflect the official policy or position of the US government.