This article appeared in Foreign Policy on January 22, 2021.

Many former U.S. intelligence officers like myself have recently been asked some variation of the question, “What should Joe Biden do to restore a broken intelligence community?”

It’s a good question. America’s intelligence apparatus has all the problems of the rest of the federal government—an aging and unrepresentative workforce, the sunk costs of expensive legacy investments, and a staid bureaucratic culture, to name a few—as well as some that are uniquely its own, such as the declining value of information in a world that’s inundated with it and the increasingly difficult proposition of keeping anything secret in an era of radical transparency. Worse, after four years of contemptuous denigration from its most important customer and having suffered the shameless politicization of its most senior leaders, Biden will inherit an intelligence community that is decidedly demoralized.

The Biden administration can’t afford to merely turn the clock back to 2016—as appealing as it may seem. Yes, having a president who regularly attends and understands his intelligence briefings and who leads sober policy deliberations will be a dramatic improvement over the alternating apathy and anarchy of the last four years. Yes, the appointment of serious intelligence professionals like Avril Haines and William J. Burns to positions of leadership is a necessary first step that will help stanch the bleeding and allow the new administration the opportunity it needs to begin correcting some of the worst mistakes of the Trump era.

But these steps are by themselves insufficient. They don’t address the larger challenges the community faces, and they won’t resolve the wicked contradiction at the heart of the intelligence community’s collector’s trap—my term for the mistaken conceit that gathering more hay will help the intelligence community find the needles it needs to understand and predict the world. Reform within the existing model won’t suffice. Biden should instead aim to remake the intelligence community, to redesign and begin building an intelligence enterprise fit for the 21st century.

As constituted, the intelligence community is a loose constellation of 18 federal agencies (now including U.S. Space Force) that are collectively responsible for the gathering, analysis, and delivery of strategic, anticipatory, and current intelligence to America’s military and political leaders. This conglomerate has, on balance, met its responsibilities remarkably well for three-quarters of a century despite its several well-known failures.

But the world that the intelligence community is tasked with helping the nation’s leaders to better understand simply no longer resembles the one it was created for. That world was slower-paced, operating at the speed of the carbon copy and the ambassador’s cable. It was less connected and thus more easily demarcated into regions—like Southwest Asia or Central America—and into topics—like terrorism or weapons of mass destruction. Power in that world was more concentrated, usually found in the hands of a relatively small number of elites who could be identified and targeted for intelligence collection. In that world, the United States squared off against the Soviet Union, and U.S. leaders’ principal challenge was a decided lack of information about that monolithic adversary—a problem the intelligence community was invented to solve.

Today, in sharp contrast, relevance is measured in hours, if not minutes. Instead of sovereign states with clearly defined borders and easily made distinctions between foreign and domestic, there are overlapping webs of influence that run over, under, and through institutions, political parties, and individuals. Issues that were once separate and distinct now brush up against and overlap one another.

Instead of being confronted by a singular superpower, today the United States is challenged by a diverse array of greater and lesser threats—some conventional, others quite novel. The threat spectrum is compounded by a parallel matrix of systemic vulnerabilities, some of which pose existential risks all by themselves. Instead of thirsting for information, today’s national security leaders are instead practically drowning in it.

Yet despite these fundamental changes in the world, the organizational and intellectual frameworks the intelligence community uses to understand it have remained remarkably consistent. Sure, there have been plenty of incremental reforms—the creation of a new agency here, the shuffling of responsibilities from one to another there. Yes, the community’s ability to gather secrets has grown dramatically thanks to a globe-spanning armature of satellites and listening posts. Yes, intelligence officers have made significant improvements to the drafting and delivery of finished analysis. But for all of these, the intelligence community still looks—and thinks—today much as it did in 1970, if not 1950.

Now it has found itself at an inflection point. Shackled to an antiquated planning process bequeathed to it by Robert McNamara in 1962, it struggles to respond quickly to the rapidly changing user requirements of the 21st century. Burdened with a broken talent management system that hasn’t been meaningfully overhauled in a generation and a glacial hiring process that discourages newcomers, it finds it much harder to recruit the next generation of intelligence officers and to retain the talented ones it already has. Intellectually, it remains trapped within an increasingly obsolete national security paradigm predicated on the notion of information scarcity, one that views the world largely in terms of threats to instead of risks from.

It must be redesigned from the bottom up, placing much greater emphasis on the bottom level. Reversing the traditional “hierarchy of privilege”—the outdated concept that spends most of the community’s resources on producing intelligence that the fewest people will ever see—it should instead deliver insight throughout the national security bureaucracy, widening the historically narrow aperture of access to facilitate understanding across the entire network rather than attempting to insert information at the top and hoping it sticks.

Instead of seeing intelligence as a product that is delivered to consumers, it should instead see intelligence as an intangible and continuous service that expands users’ mental map of the world and helps them navigate that world more easily. Rather than merely informing national security decisions through the delivery of a hard-to-find piece of information, it should empower intelligence officers to actively help their users make better ones.

The future intelligence community should be capable of continuously reconfiguring itself by adopting, combining, and discarding tools as necessary to meet changing mission needs. It should be better able to meet the greater expectations and changing preferences of a new generation of policymakers, a generation that will naturally demand more—more transparency as a matter of course but also more collaboration between what we once simplistically referred to as “producers” and “consumers.” It should be designed with distributed digital communication and collaboration tools in mind, recognizing that the most meaningful work will be done by high-performing teams composed of individuals who may not be in the same building or even in the same time zone. It should be less focused on maintaining expensive legacy technical collection platforms and more focused on recruiting and retaining creative and imaginative thinkers who give context to the information those systems collect and serve to make the complex comprehensible—which is not to say simple.

To be fair, some of the community’s leaders have recognized the imperative to change. They’ve addressed the need to modernize analytic tradecraft and developed plans to accelerate the integration of novel technologies. They have made significant steps toward the future, advocating for greater integration, more regular collaboration among analysts, and easier information-sharing among agencies.

Too often, however, their accomplishments have been achieved through workarounds that have left institutional barriers in place. Too often, parochial agencies have revealed themselves to be more concerned with protecting agency interests rather than the intelligence community’s as a whole. Some of them have resisted greater integration, made collaboration more difficult than necessary, and even flat-out refused to share data that they consider proprietary.

The pace of change must not only accelerate but broaden. The intelligence community cannot innovate simply by adopting new hardware or software as long as it retains outdated organizational and production models. It must instead be remade and must strive to differentiate between mission and tradition—that is, between things today’s users find valuable and things that are the way they are just because that’s how it has been done in the past.

The intelligence community can’t afford to rest on its laurels. Past pedigree counts for little, and previous successes guarantee nothing. As former Principal Deputy Director of National Intelligence Sue Gordon once put it, “We’ve never been as good as we are now, but we have to be better.”