The world is at an inflection point for intelligence and national security. The explosion of openly available data, paired with artificial intelligence (AI) capable of processing it at scale, is not just augmenting intelligence work—it’s redefining it. Intelligence is no longer only about secrets. It’s about using data to see clearly, decide quickly, and move first in a global technology race with profound geopolitical implications.
Today is unlike the world of 1947, the year of the CIA’s founding, when members of the intelligence community had to be hunters to find hidden secrets. Instead, data now permeates our daily existence, and intelligence can be collected globally and at scale within a digital environment. The game has changed beneath our feet. We are in the middle of renewed great-power rivalry, focused on winning the race for technologies of the future—AI, quantum computing, and synthetic biology. Mastery of these technologies will shape the geopolitical order of the next century.
History shows that paradigm shifts in technology and geopolitics demand a comparative paradigm shift in the U.S. intelligence community’s approach: from a closed, classified model to one grounded in openness, velocity, and cognition. At the center of this shift is open-source intelligence (OSINT). Powered by AI, OSINT isn’t a supplement to traditional intelligence. It is becoming the foundation.
First, OSINT must be framed as a national asset, particularly as data becomes a factor of production alongside land, labor, and capital in the AI era. Its applications cut across industry, academia, and government. Because its strength lies in accessibility and transparency, OSINT must be developed outside classified environments. That’s why it’s called “open.”
Yet many within the U.S. intelligence community still treat OSINT as secondary. That’s outdated thinking. With the vast majority of the world’s data now generated in open digital environments, OSINT must be the starting point for analysis. The question is no longer “What secrets can we uncover?” but “What does the open domain already reveal?”
Some have proposed the creation of a dedicated open-source U.S. intelligence agency. But this overlooks a key fact: The most valuable open-source work is being done outside the government. To fully realize OSINT’s potential, the United States must treat it as a national asset—not just a bureaucratic capability. That means building bridges between public and private, not more walls within the state. Instead, the private sector, where true innovation in emerging technologies such as AI lies, is far better placed to lead the United States’ future in developing scaled tools to advance the country’s intelligence and counterintelligence capabilities and doing so as colleagues operating a national asset.
Moreover, the ability of private sector intelligence providers to offer accurate, timely, and relevant insights is no longer in doubt. More than a decade ago, in 2013, the cybersecurity firm Mandiant was the first to discover the Chinese espionage group APT1. In 2016, CrowdStrike first identified Fancy Bear, the Russian hacking group. One recently retired former senior U.S. intelligence official who spoke on condition of anonymity estimated that about 95 percent of his previous work could have been obtained through OSINT.
More recently, relatively small and highly skilled teams such as those in Dow Chemical were able, for example, to harness openly and commercially available data to accurately predict Russia’s full-scale invasion of Ukraine in February 2022—ahead of many nation-states that were directly involved. Their capabilities, access to intelligence, and teams have only improved in the years since as corporate leaders deepen investment in and adoption of these tools given rapid shifts in geopolitics.
Less appreciated is a parallel revolution taking place in counterintelligence. That process, which is being equally driven by AI, has only just begun. As with intelligence more generally, the private sector, not governments, will drive forward a counterintelligence industry fit for purpose in the information era.
The U.S. intelligence community was built to counter the Soviet Union—a world of secrets, not sensors. It has evolved, but its structures and culture remain rooted in another era. Traditionally, intelligence meant collecting secret information, which was then analyzed and disseminated to decision-makers. In the past century, and even the early years of this century, intelligence required the resources of governments to pull off—to uncover the secrets that adversaries did not want discovered. And because intelligence concerned secrets, it required—almost as a matter of definition—secretive means of collection.
Today’s new digital environment offers vast amounts of information that, in the hands of the right data scientists, can be skillfully analyzed to uncover nonobvious insights. This is what the investigative reporting outfit Bellingcat has so skillfully publicly demonstrated, revealing the malign clandestine activities of the Russian government that would have previously, in the pre-digital world, been onerous tasks for Western intelligence services.
Instead of intelligence being concerned with secrets, today, it must be considered in a broader national framework: It’s about delivering advantages to decision-makers. Intelligence now is about data: who holds it; who controls it; and who can most efficiently process it, using machine learning and AI, to glean insights that enable precision in action. Whoever can achieve the latter will be the masters of this century.
If our understanding of intelligence is changing before our eyes, so too is counterintelligence. The latter can most usefully be understood as actions intended to defend against foreign intelligence offensives. One part of counterintelligence is counterespionage—catching spies, in popular parlance—but it is far broader, including activities meant to counter disinformation and even involving counterterrorism and counter-sabotage activities. Counterintelligence is one of the broadest areas of intelligence work, touching all parts of a target threat surface, whether government or private sector, for the simple reason that any part of that surface can fall prey to a hostile foreign actor.
Again, in the past century, counterintelligence was the domain of governments—it was concerned with government versus government action, nation-state versus nation-state. This is not how espionage is being waged today. Corporations stand on the front lines alongside governments in today’s tech-enabled espionage game.
To grasp the previous essence of counterintelligence, think of the nonglamorous but essential work of British intelligence officers described in the novels of John le Carré. It involved meticulous paperwork, cross-checking files, and often descended to mind-numbing activities such as scouring a foreign telephone directory in reverse to find a desired name attached to a telephone number. Within Britain’s intelligence services, there were real life officers similar to le Carré’s character Connie Sachs, who had an encyclopedic knowledge of Soviet espionage.
In the digital intelligence world unfolding before us today, Sachs would be an AI co-pilot processing real-time data flows across intelligence verticals. While the community works to incorporate AI and open-source capabilities, progress is constrained by classification, bureaucracy, and cultural resistance. Secrecy remains the coin of the realm.
To understand the precarious counterintelligence challenge facing the United States, it is necessary to appreciate how the People’s Republic of China, the principal strategic threat to the United States, understands intelligence.
China uses a whole-of-society approach to collecting intelligence, using all available means, people, and resources to steal Western military, commercial, and industrial data. It does so in fulfillment of a grand strategy, expressly enunciated by Chinese Communist Party (CCP) General Secretary Xi Jinping, intended to leapfrog ahead of the United States and become the world’s leading military and economic power.
To achieve Xi’s aim of “national rejuvenation,” the CCP blends military and civilian power and correspondingly draws little distinction between military and civilian targets in the West. Using increasingly authoritarian legislation passed under Xi, the CCP has expressly authorized its intelligence services to steal intellectual property from Western targets that cannot be developed by Chinese homegrown talent. Emerging technologies that will shape humanity this century, such as AI, quantum computing, and bioengineering, thus lie squarely within Beijing’s sights.
The CCP does not necessarily only steal secrets—it collects data of all kinds, some secret, some hiding in plain sight. This is where AI will prove to be so useful for the party—it is only through AI that the Chinese state will be able to glean insights from the vast quantity of data that it has already captured and continues to steal, storing these troves in mass data centers constructed across the country to hold domestic and foreign collected data. We can attest that the U.S.-China trade war, relaunched in April, has only intensified Chinese theft of U.S. intellectual property as the CCP tries to evade imposed trade barriers.
The Western bifurcation of public and private sectors, from an intelligence perspective, is a seam that U.S. adversaries such as China are systematically exploiting. On its own, the U.S. government cannot counter the scale of China’s intelligence offense, given its all-of-society model, which leverages gray-zone activities such as recruiting scientists or engineers inside U.S. companies to steal intellectual property. As of 2022, the FBI was opening a new China-related investigation every 12 hours. Before retiring as FBI director in January, Christopher Wray warned that China had stolen more U.S. personal and corporate data than every other country combined.
In response, a private sector counterintelligence industry is already forming as companies elevate their insider threat and security operations to tackle nation-state threats. In January, Meta CEO Mark Zuckerberg appeared on Joe Rogan’s podcast and, among other subjects, discussed the work of counterintelligence teams at Meta. Zuckerberg described their work as including identifying when platforms such as Facebook are being used by malicious actors.
It is not just Meta pursuing counterintelligence work. As of this writing, numerous job openings can be identified at leading tech companies that are, in essence, concerned with counterintelligence—at Apple, OpenAI, Meta, Amazon, Microsoft, Accenture, Crowdstrike, and Twitter/X. Every Big Tech company is adopting this model.
Unsurprisingly, the U.S. government is moving much more slowly. While the CIA, established in 1947, has undergone periodic bureaucratic updates over the years, its fundamental DNA remains broadly the same as it was nearly eight decades ago. If you were to conceive of an intelligence community fit for purpose in the 21st century, it would share many of the same objectives, but its shape would be entirely different.
In the meantime, given the pace of change and the geopolitical stakes confronting the United States, Washington should already be bolstering the tech industry’s ability to police and respond to nation state espionage.
The intelligence race is already underway—and the winners won’t be those who hoard secrets. They’ll be those who harness data, use AI, and collaborate. This is not just a transformation of tradecraft. It’s a redesign of intelligence itself. The United States should lead it.
The post The Future of Intelligence Is Open appeared first on Foreign Policy.
