A Revolution in Intelligence Affairs?
When I was a young Latin Americanist, we used to joke that Brazil was the country of the future…and always would be. So, too, talk of a revolution in intelligence affairs, somewhat akin to the ill-begotten “revolution in military affairs,” has been around for a generation. Now, it is poised to happen if—a big “if”—the intelligence agencies are agile enough.
By Gregory F. Treverton
NOTE: The views expressed here are those of the author and do not necessarily represent or reflect the views of SMA, Inc.
We have known for a long time that four pillars of the Cold War intelligence paradigm had crumbled before our eyes:
- Collection is critical. All portrayals of the canonical intelligence cycle begin with requirements, then move to collection against those requirements. In the Cold War, with one over-arching foe, that made some sense. So, given that foe’s secrecy, so did giving pride of place to intelligence’s own sources—spying, surveilling, and eavesdropping.
Neither makes sense now that information is ubiquitous. Every call for a revolution these days identifies big data, artificial intelligence (AI), and autonomous systems as drivers. By one 2017 National Geospatial-Intelligence Agency (NGA) estimate, the data available to its analysts will increase a million times over the next five years. Indeed, I worry that amidst the overwhelming information, important facts will hide in plain sight. Donald Rumsfeld famously used the distinction among known knowns, known unknowns, and unknown unknowns, but I’d add the fourth box of that matrix, unknown knowns, like those Arab men interested in flying lessons in the years before 9/11. We had forgotten, or didn’t notice. AI should at least make sure we don’t make that kind of mistake again.
The central challenge for intelligence as it makes more and more use of AI to parse through mountains of data will be to ensure that the algorithms are transparent enough, and the analysts good enough, to make sense of the results. Otherwise, they will be skeptical of what they see and unpersuasive in conveying the results to still more skeptical policymakers.
Intelligence can also learn from, and perhaps partner with, the best of the private firms doing what is still called, oddly, open-source intelligence, or OSINT. The best of those firms, like Bellingcat, form a pretty tight network in which crowdsourcing comes naturally.
- Nation-states are the focus. 9/11 upended the Cold War preoccupation with nation-states as intelligence’s targets, but it did so only to the extent of switching the focus to one set of non-state actors, namely terrorists, and now the focus once more has returned to nations, especially China and Russia. Surely, the nation-state is not about to go away, but private actors are more and more important. The Gates Foundation spends more on health in Africa than the World Health Organization. Google is probably more important geopolitically than Spain (but U.S. intelligence surely knows more about Spain than about Google). Doing better is awkward for intelligence because many of those private actors are “American,” at least when it suits them.
- So is armed conflict. In this third pillar, intelligence, like all the U.S. national security establishment, still gives pride of place to kinetic warfare. Once technology made it possible to get images from space into the cockpits of pilots nearly instantaneously, support to military operations became intelligence’s staple. When I was chairing the National Intelligence Council, I would observe how tightly integrated the Council was in Washington foreign policymaking, then add ruefully: for better or worse, you can make policy toward China without intelligence, but you can’t fight wars without intelligence.
Yet the future of conflict is not what we are best prepared for, armies on the battlefield. Rather it is in the virtual realm, groups and societies pitted against each other. Recent events, from Russia’s virtual interventions in the 2016 presidential campaign to the major hacks of 2021, have driven home that point, along with how inadequate our defenses are in that virtual realm. This observation has particular force for two intelligence disciplines:
- Counterintelligence needs to be rethought. Traditionally, counterintelligence, or CI, has been thought of as guarding against other nations penetrating our intelligence services, or permitting us to penetrate theirs. Now, in principle, the institutions to be safeguarded are those of society—they are dot.com, not just dot.gov—and the threateners are not just governments but “cyber mercenaries” available for hire but also, like those who hacked Colonial Pipeline in 2021, in it for the money.
- So does covert action. Again, U.S. covert action traditionally was thought of as secretly providing support, money, and sometimes weapons to foreign groups whose actions served U.S. interests. After 9/11, especially, the focus shifted to paramilitary operations done by the CIA in tight collaboration with Pentagon special operators, like the raid that killed Osama bin Laden. Looking to the future, covert action’s realm too will become virtual: witness the current not-so-covert debate about whether to use cyber tools to retaliate for Russia’s cyber attacks or its harboring of cyber hackers.
- Products are the product. This is the fourth pillar of intelligence that needs to be upended by the revolution. Traditionally, intelligence thought it was in the business of producing exquisite written analyses for the benefit of policymakers. Surely, it will continue to do those for a far as the eye can see, but I had one “aha” moment as NIC Vice Chair in the 1990s: we thought of our papers, National Intelligence Estimates, as our “product,” but in fact our real product was the National Intelligence Officers, people, not paper. So, too, when I later did a study of the use of the President’s Daily Brief (PDB) in the three administrations before Obama’s, senior officials liked the document but liked the briefers better. Those briefers were the opening to a conversation, to questions asked and answered.
Traditionally, too, intelligence talked of “customers” for its products. “Clients” isn’t perfect but is better, connoting a relationship between equals, not a transaction at arms’ length. President Obama received his PDB on an iPad but one stripped of all communications. Future foreign policy leaders are going to expect real iPads or better, permitting conversations with their intelligence officers 24/7. The technology is already here. What is needed is that agility by intelligence agencies in adopting it.
 Anthony Vinci, The Coming Revolution in Intelligence Affairs: How Artificial Intelligence and Autonomous Systems Will Transform Espionage, Foreign Affairs, August 31, 2020, available at www.foreignaffairs.com/articles/north-america/2020-08-31/coming-revolution-intelligence-affairs.
 These distinctions were not new with him, but he used them famously at a press briefing while Secretary of Defense, on February 12, 2002. I have thought the category he neglected to mention, unknown knowns—those things we don’t know, or remember, we know—is also critical, for instance that men from the Middle East had been interested in flying lessons before the summer of 2001.
 See Josh Kerbel and Anthony Olcott, “Synthesizing With Clients, Not Analyzing for Customers,” at www.cia.gov/library/center-for-the-study-of-intelligence/csi-publications/csi-studies/studies/vol.-54-no.-4/pdfs/Olcott-Kerbel-Client%20vs%20Customer-Extract-Annotated.pdf