We need to rethink the distinction between open sources and secrets. Too many policymakers and intelligence officers mistake
secrecy for intelligence and assume that information covertly acquired is superior to that obtained openly. Yet, the distinction
between overt and covert sources is less clear than such thinking suggests. Open sources often equal or surpass classified
information in monitoring and analyzing such pressing problems as terrorism, proliferation, and counterintelligence. Slighting
open source intelligence (OSINT) for secrets, obtained at far greater expense when available at all, is no way to run an intelligence
community. Also, we must put to rest the notion that the private sector is the preferred OSINT agent. In the end, I
would contend, the Intelligence Community (IC) needs to assign greater resources to open sources.
Mistaking Secrecy for Intelligence
Judging from their words, too many policymakers and intelligence officers mistake secrecy for intelligence. President Nixon,
for example, once belittled the Central Intelligence Agency (CIA) in words that capture the common mistake: “What use
are they? They’ve got over 40,000 people over there reading newspapers.” The president’s remarks, reflecting a persistent misperception, echo even now within the Intelligence Community. Recent
CIA recruiting literature suggests to applicants: “You can be on the sidelines, reading about global events in the newspaper.
Or you can be at the heart of world-shaping events . . . ” in the CIA. The brochure proposes a world divided between
those who read newspapers “on the sidelines” and those with access to “intelligence” within the Agency.
George Tenet, a recent director of central intelligence, was fond of defining the CIA to audiences both within and outside
the Intelligence Community with a curt phrase: “We steal secrets.” Neither from reading the CIA’s recruiting
brochure nor listening to its chief would one learn that the Agency includes an OSINT service that produces the lion’s
share of its intelligence.
Deeds also reflect the mistaken notion that secrets are all important. The Intelligence Community now includes large, well-funded
agencies for overhead imagery intelligence (IMINT), signals intelligence (SIGINT), and human intelligence (HUMINT). By all
accounts, most resources in the Intelligence Community go to such IMINT and SIGINT activities as developing reconnaissance
satellites, collecting signals, and analyzing the take. OSINT, the stepchild of the Intelligence Community, lacks its own
agency. The Foreign Broadcast Information Service (FBIS), the largest IC organization devoted to open sources, resides in
the CIA. Other OSINT units are scattered within the Department of Defense and the State Department. Alone and in the aggregate,
OSINT organizations have few people and little funding. Despite numerous surveys putting the contribution of open sources
anywhere from 35 to 95 percent of the intelligence used in the government, OSINT’s share of the overall intelligence
budget has been estimated at roughly 1 percent.
Indistinct Categories of Intelligence
Those who swear that secrets are the only true intelligence, in contrast to mere “information” found through
open means, would do well to consider the indistinct character of the categories of overt and covert in intelligence. Information
hidden behind walls of classification and special access programs may prove no more than equal in value to material available
to the public.
Overt and covert streams of intelligence are by no means completely parallel and distinct; they often mingle and meander
over one another’s territory. Covert reports at times are amalgams of press clippings. And newspaper editors, for their
part, frequently publish stories based on accurate leaks of classified material. Examples abound. Veteran CIA case officer
James Lilley learned early in his career how Chinese agents had “swindled” his office with supposedly inside information
on Chinese developments that later proved to be “embroidered versions of articles from provincial Chinese newspapers.” Similarly, European con men reportedly passed off Soviet newspaper articles as intelligence from behind the Iron Curtain
to operatives of the CIA and the West German Gehlen Organization in the 1950s. More recently, journalist Bill Gertz of the Washington Times has leaked classified information in his stories. His
published photocopies of actual intelligence documents underscore how the overt and covert streams mingle.
The more one considers the problem, the less distinct appears the distinction between open information and secrets. Let
us consider the case of the B-29 bomber aircraft, whose use in the Second World War was reportedly classified. Samuel Halpern,
an officer of the Office of Strategic Services (OSS), recalled how he once surprised an admiral by referring in his briefing
to the B-29 Super Fortress bombers. When the admiral demanded to know how Halpern knew of the “highly classified”
aircraft, the OSS officer replied that he had learned of the bomber through monitored Japanese radio broadcasts. In short, what is classified to some is open information to others. This can lead to the absurd situation where foreigners
learn details of US intelligence operations in their country through their national media, while the American public and Intelligence
Community remain unaware of the overseas exposure. Perhaps “unilateral secret” would be the proper term for this
The Value of Open Sources
Not only are open sources at times indistinguishable from secrets, but OSINT often surpasses classified information in
value for following and analyzing intelligence issues. By value, I am thinking in terms of speed, quantity, quality, clarity,
ease of use, and cost.
Speed: When a crisis erupts in some distant part of the globe, in an area where established intelligence
assets are thin, intelligence analysts and policymakers alike will often turn first to the television set and Internet.
Quantity: There are far more bloggers, journalists, pundits, television reporters, and think-tankers in the
world than there are case officers. While two or three of the latter may, with good agents, beat the legions of open reporters
by their access to secrets, the odds are good that the composite bits of information assembled from the many can often approach,
match, or even surpass the classified reporting of the few.
Quality: As noted above, duped intelligence officers at times produce reports based on newspaper clippings
and agent fabrications. Such reports are inferior to open sources untainted by agent lies.
Clarity: An analyst or policymaker often finds even accurate HUMINT a problem. For example, when an officer
of the CIA’s Directorate of Intelligence (DI), reads a report on a foreign leader based on “a source of unproven
reliability,” or words to that effect, the dilemma is clear. Yet, the problem remains with a report from a “reliable
source.” Who is that? The leader’s defense minister? The defense minister’s brother? The mistress of the
defense minister’s brother’s cousin? The DI analyst will likely never know, for officers of the Directorate of
Operations (DO) closely guard their sources and methods. This lack of clarity reportedly contributed, for example, to the
Iraqi WMD debacle in 2002-03. The DO reportedly described a single source in various ways, which may have misled DI analysts
into believing that they had a strong case built on multiple sources for the existence of Iraqi weapons of mass destruction. With open information, sources are often unclear. With secrets, they almost always are.
Ease of use: Secrets, hidden behind classifications, compartments, and special access programs, are difficult
to share with policymakers and even fellow intelligence officers. All officials may read OSINT.
Cost: A reconnaissance satellite, developed, launched, and maintained at a cost of billions of dollars, can
provide images of a weapons factory’s roof or a submarine’s hull. A foreign magazine, with an annual subscription
cost of $100, may include photographs of that factory’s floor or that submarine’s interior.
Beyond this general argument for open sources, I would maintain that OSINT often equals or surpasses secrets in addressing
such intelligence challenges of our day as proliferation, terrorism, and counterintelligence. When a nation develops a weapon
of mass destruction, for example, hundreds or even thousands of engineers, scientists, and manufacturers may join the program.
Bureaucrats and traders may sell the weapons abroad. The OSINT target is immense. Engineers attend conferences; scientists
publish scholarly articles; manufacturers build production lines; bureaucrats issue guidelines; and traders print brochures
for prospective clients. Many paper trails wind around the world beyond whatever may surface in the media.
Before terrorists act, they issue warnings, religious leaders of their community deliver sermons, and political leaders
plead their cause. Open sources, while they may not tell us where the next bomb will explode, do allow us to understand the
terrorist agenda and act thereby to address grievances or launch competing campaigns for hearts and minds.
When foreigners seek to tap US technology abroad or on our soil in order to evade embargoes or leapfrog the R&D process
at our expense, open sources may alert counterintelligence officers to their activities. For example, the National Counterintelligence
Executive has published reports based on Korean media from both sides of the DMZ to bring to light North and South Korean
efforts to acquire Western technology both abroad and in the United States.
The Cost of Slighting Open Sources
Arguing that we need to rethink the distinction between open information and secrets, which is more blurred than many think,
and that OSINT is often more useful in addressing intelligence challenges, I would further maintain that Washington’s
slighting of open sources is no way to run an intelligence community. In earmarking only one of every hundred dollars in the
intelligence budget and assigning some similarly meager percentage of IC personnel to OSINT, our policymakers and intelligence
executives are learning less than possible about our nation’s challenges while paying a higher price than necessary.
DO officers without access to foreign media published uncounted numbers of bogus reports based on Chinese, Soviet, and other
newspaper articles. We are also almost certainly spending large sums today to obtain covertly information similar or
identical to that openly available. Rather than learn through HUMINT or SIGINT that a scientist of interest attended an international
conference, for example, would it not be better simply to acquire, then print or report the contents of the conference proceedings?
Open acquisition would likely be less expensive, and all policymakers and analysts would have access to the information.
Policymakers and intelligence executives would also do well to resist the siren call of those who argue that we should
simply privatize OSINT. Private corporations are an excellent source of dictionaries, software, and contractors for our government.
But private companies alone are no substitute for accountable, dedicated OSINT professionals in government offices. Let us take the vital issue of translation as an example. Contractors— whether individuals, translation agencies, or
research companies (the latter generally subcontracting with translation agencies or independent translators for the talent
they lack in house)—today translate most of the foreign newspapers, scientific journals, and other open information
for the Intelligence Community. They do so under the lead of cleared OSINT officers who, knowing both the requirements of
the Intelligence Community and the mysteries of the foreign media, manage the translation flows to provide answers to intelligence
questions. Staff officers are also available to translate priority items themselves on a crash basis when contractors are
unavailable. Staff officers serve one master. Contractors, busy with a mix of assignments from corporate and government customers,
often are unavailable when most needed.
Ideally, in my view, the government should develop its own sizeable cadre of translators. Yet, that would be much more
expensive than the present system. Some would argue for the opposite path of privatizing OSINT, which would mean intelligence
analysts, case officers, and others buying their translations directly from the private sector without OSINT officers to apply
their general requests against the appropriate media for the right information or to edit, often heavily, contractor translations
that are frequently of poor quality.
The above logic, which applies to media analysis, targeting, and other OSINT functions as well as to translation, suggests
that the government should retain its OSINT capabilities. The Intelligence Community requires staff officers to lead the contractors.
To use an analogy from history, private corporations may have supplied the aircraft, landing craft, and rifles for D-Day,
but General Eisenhower, his military staff, and soldiers in uniform took the beaches at Normandy.
Assigning Greater Resources
I have maintained that (1) secrets are not identical to intelligence; (2) the distinction between overt and covert sources
is more blurred than commonly imagined; (3) open information often equals or surpasses classified material; (4) slighting
OSINT is no way to run an intelligence community; and (5) the private sector is no substitute for the government in applying
open sources to address today’s intelligence challenges. I can only conclude that Washington needs to assign greater
resources to open sources. Whether we create a national OSINT center or leave FBIS and its counterparts right where they are
is less important than the issue of dollars and people. Putting all the meager OSINT offices together in a single center,
without added funding, would be analogous to a poor man combining several small bank accounts into one—he would still
be poor. With greater resources, perhaps a doubling of OSINT spending to roughly 2 percent of the intelligence budget, we
would see an impressive increase in intelligence available to all in government. It would even permit covert collectors to
focus with greater precision on areas truly beyond the reach of open sources.
Charles E. Lathrop, The Literary Spy: The Ultimate Source for Quotations on Espionage & Intelligence (New Haven:
Yale University Press, 2004).
I came across the brochure, Challenges for a Changing World, while on a college recruiting trip in 2003. For an example
of DCI Tenet’s fondness for the phrase “We steal secrets,” see an account of his 1999 speech at his alma
mater, Cardozo High School in Queens, New York, in Vernon Loeb’s “Back Channels” column, The Washington
Post, 18 June 1999. In my use of the word intelligence, I mean vetted information from any source. For more on how I define
intelligence, see my article “Sailing the Sea of OSINT in the Information Age,” Studies in Intelligence
48, no. 3 (2004): 50–51.
The National Reconnaissance Office (NRO) develops reconnaissance satellites. The National Geospatial-Intelligence Agency (NGA)
provides IMINT. The National Security Agency (NSA) handles SIGINT. The CIA devotes most of its resources, apart from its analytical
directorate, to collecting HUMINT and developing technical support for such operations. FBIS, established in 1941
in the Federal Communications Commission, exists today as an OSINT service within a HUMINT organization. For OSINT’s
contribution and its funding, see Joseph Markowitz, “The Open Source Role,” Horizons 1, 2 (Summer 1997):
James Lilley, with Jeffrey Lilley, China Hands: Nine Decades of Adventure, Espionage, and Diplomacy in Asia (New
York: Public Affairs, 2004), 82–83.
For reference to the European cons and an earlier account by Ambassador Lilley of the China debacle, see Evan Thomas, The
Very Best Men: Four Who Dared: The Early Years of the CIA (New York: Simon & Schuster, 1995), 154.
For an example of Gertz’s use of leaks, see the copies of classified documents in the appendix to his book Betrayal:
How the Clinton Administration Undermined American Security (Washington, DC: Regnery, 1999). The appendix, titled
“The Paper Trail,” covers pages 217–84 of a book running 291 pages. In other words, his classified appendix
accounts for nearly one page of every four in the book.
Samuel Halpern, “Remembering 109,” OSS Society (Fall 2001), 5; accessed at www.oss.org.
While the US media has no habit of exposing CIA and DoD actions and organizations in East Asia, to pick a region, the
same cannot be said of the Asian media. Pyongyang’s press, for example, publishes articles detailing the alleged flights
each month of US reconnaissance aircraft against North Korea. Seoul reporters, too, engage in the sport of “spot the
spook” by telling Korean readers that this or that office of the US Embassy is a CIA station. Even the media in Japan,
our bedrock partner in the Pacific, repeatedly treat the public to stories of US officers under non-official cover and other
The image of CIA officers tuning in CNN to watch the breaching of the Berlin Wall underscores this point. See my article,
“Sailing the Sea of OSINT in the Information Age,” Studies in Intelligence 48, no. 3 (2004): 47.
Walter Pincus, “CIA Alters Policy After Iraq Lapses,” The Washington Post, 12 February 2004: A1.
For a report from the National Counterintelligence Executive on Pyongyang mobilizing overseas Koreans to acquire foreign
S&T information, see “Overseas Koreans Contributing Technical Literature to DPRK,” Counterintelligence
News & Developments, September 2001. For a story about South Korean plans to build a biotechnology facility in California,
see “South Korea: Biotech Consortium To Build Tech-Transfer Facility in San Diego,” News & Developments,
January 2002. Both reports are available on line at www.ncix.gov.
I do not maintain that contractors are a different species from intelligence officers. Indeed, they are often our retired
colleagues who have traded staff badges for contractor badges.
Stephen C. Mercado is an analyst in the CIA Directorate of Science and Technology