22 November 2002

But I also want to emphasize that I would not have these views had I not come to them from the vantage point of my 20 years in the intelligence business, and particularly the last seven with the Foreign Denial and Deception Committee.  This committee represents a small interagency effort of the Intelligence Community to understand how foreign adversaries—hostile countries and terrorist groups, for example—learn about, then try to defeat, our secret intelligence collection activities.  I have come to appreciate, perhaps more than most who do not or cannot see this problem from the unique vantage point I have, that unauthorized disclosures of classified intelligence pose a serious, seemingly intractable, problem for US national security.  The Director of Central Intelligence, George Tenet, made the point in an interview that unauthorized disclosures of classified information “have become one of the biggest threats to the survival of US Intelligence.”  (USA Today, 11 October 2000).  A skeptical public can rightly question whether the DCI isn’t exaggerating the seriousness of the problem.  Unfortunately, he is not, and no intelligence specialist with real familiarity with this issue would disagree.

This presents an important anomaly in public discourse:  Nearly all of the compelling evidence in support of this argument is available only in the classified domain.  It thus seems daunting to make a persuasive public case for new laws to address unauthorized disclosures when so little of the evidence for it can be discussed publicly.  So proponents for better laws—you will soon see why I am one of these—sometimes feel that this isn’t a fair fight.  Freedom-of-the press advocates and professional journalists who control the press therefore exert disproportionate influence on this debate, at least when compared to advocates of criminal penalties, as I am here, for the leaking and publishing of sensitive classified intelligence.  But I have come to believe that First Amendment objections to civil and criminal penalties for disclosing classified intelligence are probably exaggerated.  And that once we get over this hurdle, it’ll be more of a fair fight, a more reasonable debate.  It is in this spirit that I offer the following remarks.

I have four basic propositions:

1. The problem of unauthorized disclosures of classified intelligence is so serious that this issue demands action from Congress as well as the Executive Branch.
2. An important reason for the seriousness of the problem lies in poor laws and poor law enforcement.
3. The remedies for this malady are not all that dramatic and constitutionally bone-jarring, but are rather found in a sensible combination of new laws, amending old ones, and better enforcement.
4. The consequences of legal inaction are high—perhaps higher than we should ask the American citizenry to bear.

The scope of my concern with classified information here extends only to intelligence, which encompasses intelligence information, activities, operations, sources, and methods.  I exclude from my purview other kinds of classified information such as military (war plans, weapons systems, etc.) and diplomatic secrets, not because they are unimportant, but because I believe that intelligence increasingly requires a distinct legal identity.    

I.  The Seriousness of Unauthorized Disclosures

“It is ‘obvious and inarguable’ that no governmental interest is more compelling than the security of the Nation.”-- US Supreme Court in Haig v. Agee (1981)

It is a myth, too commonly held outside the Intelligence Community, that leaks really don’t do much harm.  The genealogy of this erroneous view traces to the publication of the Pentagon Papers.  After much government carping about all the damage that those Top Secret revelations in the press would do to US national security if disclosed to the public, few today would claim that any damage was done at all.  And I am unaware of any that was done to intelligence.  The Pentagon Papers flap took us off the scent.  The view that leaks are harmless is also further nourished by another popular myth that the government over-classifies everything—including intelligence—and classifies way too much.  This seduction has become a creed among anti-secrecy proponents.  But this too, at least in intelligence, I would argue, is wrong. 

A recent classified study of media leaks has convincingly shown that leaks do cause a great deal of harm to intelligence effectiveness against priority national security issues including terrorism.  This is principally because the press has become a major source for sensitive information for our adversaries on US intelligence—its information, activities, operations, sources, and methods.  Unfortunately, there are many serious leaks of US intelligence that cumulatively provide substantial information to foreign adversaries.  For example, at CIA alone, since October 1995, there have been hundreds of investigations of potential media leaks, and a significant number of these were referred to the Department of Justice for follow-up actions.  Leaks that have damaged the National Security Agency’s (NSA) signals intelligence sources and methods also number in the hundreds in recent years; and dozens of these cases have been referred to Justice.  The National Imagery and Mapping Agency (NIMA) too has experienced roughly a hundred leaks just since 2000 that have damaged US imagery collection effectiveness. 

It is impossible to measure the damage done to US intelligence through these leaks, but knowledgeable specialists assess the cumulative impact as rather significant.  Some losses are permanent and irreversible; others can be recovered, though sometimes only partially, but only at additional expenditure of substantial resources that could well be spent elsewhere. 

While leaks of classified information are often intended to influence or inform US audiences, foreign intelligence services and terrorists are close and voracious readers of the US press.  They are keenly alert to revelations of US classified information.  For example, a former Russian military intelligence (GRU) officer wrote:

“I was amazed—and Moscow was very appreciative—at how many times I found very sensitive information in American newspapers.  In my view, Americans tend to care more about scooping their competition than about national security, which made my job easier.” 

(Stanislav Lunev, Through the Eyes of the Enemy, Washington, D.C., 1998, p. 135).     

I will call this the Lunev Axiom:  Classified intelligence disclosed in the press is the effective equivalent of intelligence gathered through foreign espionage.  What we need to understand are the legal implications of this key principle.  (Some of its historical effects are summarized in the text box below).  Importantly, more than just Russian intelligence officers understand this.  Key adversaries of the United States such as China and Al Qa’ida derive a significant amount of their information on the United States, including on US intelligence, from the open-source media, including the Internet.

         How press leaks hurt US intelligence.  The Intelligence Community faces improved foreign countermeasures as adversaries use leaks to expand their understanding of US intelligence, such as imaging capabilities, for example.  Press disclosures have revealed to our adversaries—directly and indirectly—unique elements that underpin our analytic tradecraft.  Thoughtful manipulation by adversaries, as well as friends, of such knowledge exposed in the press—as happened before the Intelligence Community missed the surprise Indian nuclear tests in May 1998 (where official demarches also piled on the damaging effects of leaks)—threatens our ability to provide policymakers with a decisive information advantage.

In addition, effective intelligence depends on cooperative relationships with friendly governments and with individuals who trust the United States to protect their confidences.  Press disclosures can undermine these relationships, making both governments and individuals reluctant to share, thereby inhibiting intelligence support crucial to informed policymaking.  For example, a 1998 newspaper report provided lengthy coverage regarding UNSCOM, the UN Special Commission charged with inspecting Iraq’s WMD facilities following the Gulf War.  The reports were cited in subsequent worldwide media reporting.  Although the articles contained many inaccuracies, the information contained in them interfered with the government’s ability to aggressively pursue a long-standing US policy on Iraqi weapons inspections.  Importantly, unauthorized disclosures clearly have degraded the United States’ ability to obtain intelligence on Iraq.  Press disclosures revealing imagery-derived military intelligence on Iraq have included movement of missile systems, construction of a new command and control network, and, following the September 11th terrorist attacks, dispersal of equipment.

More recently, as Congress continues its investigation of whether September 11th??? was an intelligence failure, it should come to notice—at least I hope it does—thats important intelligence collection capabilities against Osama Bin Ladin and Al Qa’ida were lost in the several year period leading up to 9/11.  With the concurrence of NSA, the White House officially released just one of these.  As press spokesman Ari Fleischer explained: 

“And let me give you a specific example why, in our democracy and in our open system, it is vital that certain information remain secret.  In 1998, for example, as a result of an inappropriate leak of NSA information, it was revealed about NSA being able to listen to Osama Bin Ladin on his satellite phone.  As a result of the disclosure, he stopped using it.  As a result of the public disclosure, the United States was denied the opportunity to monitor and gain information that could have been very valuable for protecting our country.”  (White House press statement, 20 June 2002).

Now, is there any connection between law enforcement and unauthorized disclosures?  If this were a statistical problem, we could probably show a highly inverse correlation:  The less the law enforcement, the greater the leaks of classified information—and probably the other way around as well.  But since there has been only a single example of a prosecution for an intelligence leak, Navy analyst Samuel Loring Morison in 1985, a statistical approach is absurd.  Let us just note that the apparent absence of civil or criminal penalties for leaking and publishing classified intelligence establishes a law enforcement climate of utter indifference—actually permissive neglect.  The unofficial message seems to be:  Leak all you want, and no matter how much, or how serious, nothing will happen to you. 

Actually, for perpetrators, rather than penalties there seem only benefits to leaking.  Anonymous government officials who do it can skew public debate in their favor by selectively leaking intelligence that supports their policy positions.  And journalists and book publishers can also gain policy influence, brandishing relevant intelligence that their opponents may not have seen, and cannot easily refute.  But also, over time, they can also gain public renown and recognition; better newspaper, magazine, and book sales; as well as bigger incomes and profits merely by exploiting the classified materials that law-breaking government officials provide to them.  It is an unholy alliance that works exceedingly well as long as the legal climate remains indifferent to it.

   Is leaking classified intelligence against the law?  Probably—but it depends on who you ask.  And you wouldn’t know it from the prosecutions data:  Only Morison alone might say so—and he was pardoned by Bill Clinton (who also vetoed the “Shelby Amendment” anti-leaks law in the FY01 Intelligence Authorization Act), as our past President was just leaving office last year.

      It is precisely the legal ambiguity of leaking that is the heart of the problem here.  Certainly there are laws against it—chiefly the 1917 espionage law Title 18 USC, Sections 793 (d)-(e), and 798; and the narrower Title 50, Section 421.  One could devote a whole legal seminar to what is wrong with these laws—I urge this committee to consider this—but suffice it to say here, in my layman’s view, that a law that is almost never enforced is either unneeded or useless.  I contend that anti-leaks laws are seriously needed—but since the present ones are virtually unenforceable, they are useless.

This problem is not new.  Let me quote from the now historic “Willard Report” (after its chairman Richard K. Willard, Deputy Assistant Attorney General, Report of the Interdepartmental Group on Unauthorized Disclosures of Classified Information, 31 March, 1982, prepared for the President).  It concludes: 

“In summary, past experience with leak investigations has been largely unsuccessful and uniformly frustrating for all concerned . . . .  This whole system has been so ineffectual as to perpetuate the notion that the government can do nothing to stop the leaks.”  (Emphasis added).

The recommendations of the Willard Report for legal correctives resulted in proposed legislation in 1984.  Although supported by OMB and the Administration, the Intelligence Community later withdrew the recommended draft legislation due to a perceived lack of support. 

Twelve years later, responding to a request from the Assistant to the President for National Security Affairs, the National Counterintelligence Policy Board completed another study and reported no discernible change.  (Report to the NSC on Unauthorized Media Leak Disclosures, March 1996).  The report explained the continuing failure as a result of two key factors:

1. A lack of political will to deal firmly and consistently with unauthorized executive branch and Congressional leakers.  Unsuccessful government efforts to prevent unauthorized disclosures were attributed largely to a lack of political support.
2. The use of unauthorized disclosures as a vehicle to influence policy.  [Citing the 1987 Tower Commission Iran/Contra report]:  ‘Selective leaking has evolved to the point that it is a principal means of waging bureaucratic warfare and a primary tool in the process of policy formulation and development in Washington’.”   (Boldfaced emphasis in the original)

Given the palpable history of failure in protecting classified intelligence information from press disclosures—and given their epidemic proportions and the deleterious consequences they wreak in countermeasures to US collection effectiveness—it is fair to question why past failed approaches should be expected to work today.

There has never been a general criminal penalty expressly for the unauthorized disclosure of classified intelligence to the press.   Although intelligence leaks can technically be prosecuted under espionage statutes cited above (18 USC 793 and 798), only a single case ever has (US v. Morison).  Given that literally thousands of press leaks have occurred in recent years—many serious and virtually all without penalty—it is clear that current laws do not provide an effective deterrent to leakers or to journalists and their media outlets that knowingly publish classified intelligence.   

This brings me to my central thesis that bad laws are hard to enforce.  A penetrating critique of what passes for anti-leaks laws is provided in Ballou and McSlarrow, a comprehensive Note in the Virginia Law Review, June 1985.  Although written before the Morison prosecution, the chief points remain as valid today as when written.  Let me cite a key passage that highlights the responsibility of Congress: 

Let me turn here to some proposed remedies.  I defer to the drafting skills of competent attorneys to translate any promising ideas here into workable legislation.  I group my suggestions into three categories:  Write new law.  Amend old ones.  And enforce them all—new or old.

1.  New laws.  Given the fact that many thousands of leaks of classified intelligence in recent years have severely damaged intelligence effectiveness, thereby jeopardizing the nation’s security—and existing penalties provide no effective deterrent to leaking—we urgently need a comprehensive anti-leaks statute to empower law enforcement and investigators to better protect intelligence.  What attributes should such a new law have?  Above all, a new law should reduce confusion and add clarity to what is legally permissible and what is not.  And it should deter and punish leaks of classified intelligence.  Specifically, it should:

• Unambiguously criminalize unauthorized disclosures of classified intelligence. 
• Hold government leakers accountable for providing classified intelligence to persons who do not have authorized access to that information, irrespective of intent or willfulness; and hold unauthorized recipients of that information accountable for publishing it. 
• Define intelligence information, including substantive content, activities, operations, sources and methods distinctly from defense information, creating a distinct protected category for intelligence that does not require meeting an unneeded burden of “defense information.”
• Provide better protection to especially sensitive and highly classified intelligence information in trials and other judicial proceedings than is presently afforded through the Classified Information Procedures Act (CIPA).
• Revisit the clean drafting guidelines proposed by Willard 20 years ago (as the Shelby Amendment appears to have done) that simplify the elements describing unauthorized disclosures of classified information. 

Most of the above attributes can be accomplished without inviting serious debate over First Amendment issues.  Some invite more controversy. 

In addition, a separate new law that we should consider would constitute a technical counterpart to the Intelligence Identities Protection Act (50 USC 421), crafted to provide the same protection to technical sensors deployed on any platform (space, air, land, sea) that is now afforded to human operations. 

Should journalists have legal accountability? Absolutely, in my view.  All agree that the first line of enforcement must be drawn at offending government officials who unlawfully steal and disclose classified information.  And, like citizens everywhere, government officers too have different opinions on the propriety of holding journalists legally liable for publishing classified intelligence.  Still, I believe that to be fully effective, a worthy law should also hold uncleared publicists—i.e., journalists, writers, publishing companies, media networks, and websites that traffic in classified information—accountable for such disclosures.  Specifically, media representatives should be held responsible for publicizing—thereby to terrorists and other US adversaries—significant intelligence information they know to be classified, and whose public exposure threatens to reduce US intelligence effectiveness by damaging valuable sources and methods.  They should also be held liable for unlawful possession of classified documents when they have them.

Legal accountability for journalists is necessary because declassification authority is assigned by law exclusively to government officials, elected and appointed, through lawful procedures.  Journalists who publish classified intelligence arrogate to themselves an authority legally vested in government that they do not by right possess.  In publishing classified intelligence, no journalist can convincingly claim the constitutional right to do so.  Any journalist’s First Amendment right to publish information does not appear to—and should not—extend to disclosing lawfully classified intelligence information.  In any case, this legal claim remains to be made or adjudicated. 

Arguably, a close reading of 18 USC 798 (sometimes referred to as the SIGINT statute) and surely 50 USC 451 (Intelligence Identities Protection Act), will show that journalists are already legally accountable for publishing leaked classified intelligence.  But these statutes (too narrowly drawn, and considerable intelligence escapes their purview) are apparently unenforced to date, and remain to be tested in the courts. 

Like government officials, journalists also exercise a public trust.  But they exercise it without any apparent legal accountability for violating the public trust when they reveal the nation’s secrets.  This is wrong.  Legal accountability for journalists is especially needed in the absence of an enforceable code of ethics for journalist conduct.  While the overwhelming bulk of journalists recognize the ethical implications of compromising sensitive intelligence sources and methods, a few egregious offenders do not, and their complete lack of accountability for costly intelligence compromises that jeopardize the nation’s security must change under the force of law. 

First Amendment issues.  Constitutional experts will address First Amendment implications of any proposed legislation that may be interpreted to constrain freedom of the press.  For the most part, I believe that holding publishers of classified intelligence legally accountable should not necessarily invite constitutional apprehensions.  Constitutional arguments will have to consider the following:

• The government’s executive authority to classify information is firmly established in law.
• Congress’s willingness to regulate publications where the potential for serious harm exists is already established in the Intelligence Identities Protection Act (50 USC 421; see Ballou and McSlarrow, p. 7), and probably in 18 USC 798 as well. 
• One leaker has been convicted of providing classified information to the press, and this decision was upheld on appeal (US v. Morison, 844 F.2d 1057, 4th Cir., cert denied, 488 US 908, 1988).
• Publishing classified intelligence has not been demonstrated as a constitutionally-protected right.
• A compelling argument for extending the “harm principle” (below) to protecting classified intelligence from press exposure can be made when the nation’s security can be shown to be in jeopardy as a consequence, for example, of the media’s assistance (unwitting, to be sure) to US adversaries such as the terrorists who planned and conducted the attacks in New York and Washington on September 11th. 

Of course, the inherent tension between First Amendment rights and the government’s interest in protecting national security is dynamic, and may never be solved “once and for all.”  But the current balance so favors First Amendment rights that compelling constitutional interests involving national security can be superseded.  This seems certainly the case with unauthorized media disclosures that result in the compromise of intelligence effectiveness and thus jeopardize national security.  Here we should entertain redressing a potential legal imbalance by reconsidering a time-tested democratic principle enunciated by the preeminent philosopher of liberty, John Stuart Mill:

“ . . . the only purpose for which power can rightfully be exercised over any member of a civilized community, against his will, is to prevent harm to others.”  On Liberty (1859).

Under the “harm principle”—for example, yelling “FIRE!” in a theater when there is none—a variety of exceptions to free speech are well established in American law such as obscenity, defamation, breach of peace, and “fighting words.”  To this list we should add the compromise of US intelligence information, activities, operations, sources, and methods, required in the service of the Nation’s security. 

      2.  Amend existing law.  Addressing the concern raised in the report of the National Counterintelligence Policy Board, if we lack the political will to write a new law—and I am convinced that lack of will is our chief obstacle here—then I urge that we amend our present, defective, laws to help us curtail the loss of present and future US intelligence capabilities.  Here is what I would do: 

• Amend 18 USC 793:

-- To establish a distinct legal identity for intelligence information, activities, operations, sources and methods—apart from national defense, since considerable intelligence activities can be argued as unconnected with national defense, thus removing an additional prosecutorial burden to satisfy.

-- To remove the burden of intent or “willfulness” standards, requiring only that the government show (i) that classified intelligence information was leaked, and (ii) that this leak could damage US intelligence collection effectiveness.  I would restrict any “intent” burden only to establishing a leaker’s intent to disclose classified intelligence instead of the higher culpability bar establishing intended damage to the nation. 

• Amend the Intelligence Identities Protection Act, 50 USC 421:

-- To remove the burden of establishing “patterns” of disclosures, since some singular disclosures are so serious, perhaps resulting in loss of life, that legal penalties for exposing sensitive agents who risk their lives to help the United States and its allies must be clearly established. 

-- To remove the intent standard because agent identities can be revealed to discerning readers (such as foreign intelligence services or terrorists organizations) through merely descriptive information even when actual names are withheld.

-- And, unless we craft a new law to accomplish this, I would broaden the scope of this narrow statute that now covers only human operations to also apply to technical collection activity as well, including from spaceborne sensors.

• Amend 18 USC 794 to include terrorist organizations and groups, along with “foreign governments or agents thereof” as is currently written.
• Amend CIPA to afford much greater protection during investigative and judicial proceedings for highly compartmented information, which, when leaked, may not even be investigated or officially reported for prosecution owing to an understandable government incentive to not call further attention to a particularly sensitive activity or capability. 

3.  Enforce our present laws.  Perhaps even more than writing new laws or amending old ones, enforcement will require real political will—surely more than we’ve seen since Morison.   What would I do?

First, acknowledge the Lunev Axiom:  Recognize that government leakers and the journalists who publish the classified materials they provide do the equivalent work of spies.  Even if their motives differ, the effects are often the same.  Through press leaks, unauthorized disclosures can be every bit as damaging as espionage because of the focused exploitation of the US press by terrorists and hostile governments.  If leakers and journalists were caught providing some of this classified information clandestinely to a foreign power, they could be prosecuted for espionage.  But if they publish in the press—where their leaked sensitive information becomes available to all governments and terrorists, not just one—they derive effective immunity from prosecution under a government that lacks the will to enforce its laws.  Let me state this categorically: Adversarial foreign countries and terrorists rely heavily on the US press to glean sensitive information about intelligence in order to deploy countermeasures against it.  Since such disclosures can have the same effect as espionage, we should treat leakers to their collaborating journalists as subject to the same laws that apply to spies whose work is more clandestine, but sometimes no more damaging.  We should:

• Enforce 18 USC 794 against providers and publishers of classified information whose disclosures are thus made available to hostile foreign governments just as surely as any spies’ disclosures.
• Subpoena, in the course of legal proceedings to recover stolen government property, classified intelligence documents that we believe are in the possession of government leakers or journalists, and thus outside of the normal physical protections that the US government provides to sensitive classified intelligence information.
• Prosecute government officials, journalists and publishers who are found to be in possession of documentary classified intelligence for possession of stolen government property under 18 USC 641.

We should, in sum, recognize that sensitive intelligence information is classified by this government for good reasons—precisely because its protection really is essential to the security of the nation.  But the legal protections we afford it are woefully insufficient, and not nearly as good as those we provide to other government or government-protected information—such as crop estimates, certain economic data, and insider trading for securities, for example—whose acquisition by foreign adversaries and terrorists wouldn’t make any difference at all. 

IV.  Consequences of Failure to Act

“If the law supposed that,” said Mr. Bumble, “the law is an ass.”-- Charles Dickens, Oliver Twist

Finally, why have unauthorized disclosures become a public policy issue now?  Because years of inaction, indifference, and permissive neglect are taking an enormous toll.  And the toll is higher still since 9/11.  Intelligence leaks do serious and often irreversible damage to our sensitive collection capabilities.  By publicly unveiling unique US collection capabilities through leaks, the media actively help our adversaries to weaken US intelligence.  These disclosures offer valuable insights—at no cost to adversaries—into possible errors in their assessments of US intelligence effectiveness, as well as useful feedback on how well they succeed or fail in countering US intelligence.  This feedback also increases the risk of manipulation of our intelligence.  Leaks offer a bonanza of intelligence information for hostile states and terrorist organizations.  When added to other sources of readily available intelligence, the aggregate effect of leaks is to provide a comprehensive picture of our intelligence capabilities—and vulnerabilities—to our adversaries, facilitating countermeasures to them, thereby damaging US national security.

Unless comprehensive measures are taken to identify and hold leakers and their publishing collaborators accountable for the significant, often irreversible, damage they inflict on vital US intelligence capabilities—and, by implication, the degraded support such weakened intelligence offers to policymakers and warfighters—the damage will continue unabated.  Conceivably, without some legally effective corrective action, the situation could even worsen.   Under this scenario:

• Policymakers, warfighters, and military planners should expect that their intelligence on significant national security issues will be less accurate, complete, and timely than it would be without foreign countermeasures made possible by unauthorized disclosures.   
• American citizens should expect that timely warning of surprise attacks against the United States by terrorists or other hostile adversaries will be degraded because key collection activities have been rendered less effective through unbridled leaks. 
• US taxpayers should expect that their multi-billion dollar collection programs will be less cost-effective than they would otherwise be if foreign adversaries were not learning how to neutralize such programs through extensive classified information readily available in open sources as a result of unauthorized disclosures