The ‘No Classified Information’ State: An Open Source Solution to a National Security Problem.
0. Abstract
Could a state with no secrets function better when protecting national security than a state that keeps information away from the general public? In this brief article, we will inspect the reasons for keeping classified information, what they are meant to protect and how they protect national security. We will present the method used by Israel, which is similar to most states. Israel’s approach, which is to keep all the information from the public, failed in general and caused nothing but costs on privacy, freedom of expression and national budgets.
Following our review, we will compare the classified information model to a model in information security, called Security through Obscurity and present how this model was perceived as flawed. Against it, we will present the Open Source Model, which creates transparency towards the general public, allowing it to inspect the security flaws, and therefore creates stronger protection.
Our conclusion would be that better national security could be reached by removing all classified information and disclosing all information to the general public. We believe that by making the information public, the cost of the censorship apparatus will be eliminated. We also believe that by adopting a ‘no classified information’ approach, governments may improve physical security when they rely on the foundations of open source security as detailed herein.
In my brief argumentation I will use the Israeli law, but provide some examples from other cases.
1. Classified Information and what it Protects.
Every state has its secrets. States choose, in certain cases to classify information from the general public. Classifying information goes back as far as Greek times, and goes under the standard four categories: Top Secret, Secret, Confidential and Restricted. Israel has four apparatuses which are in charge of Confidential information: The Information Security Department, whose goal is to prevent classified information from leaking from the army, The Military Censorship, which operates under the Defense Ordinance (Time of Emergency), 1945, that controls media publication and telecommunication, and has authority to refuse the publication of any information that has any relation to national security, the General Security Service (Shin Bet) that acts according to the General Security Service Act of 2002, where clause 7(2) allows the service to classify documents and determine how to handle such documents and the Director of Security of the Defense Establishment, which is in charge of security in military industries, research facilities and other national security industries.
Some authorities in classifying information do not appear to exist in laws, and some operate under the vague and broad exemption added in the Freedom of Information Act, 1998. Clause 9 to the Israeli FOIA exempts disclosure of any information which may harm national security, foreign relations, public safety or a person’s well-being. Even in cases where classified information was disclosed, the courts still allowed the security agencies broad discretion as to what to blur out (HCJ 258/07 Zehava Galon v. The Governmental Committee for Inspecting the Battles in Lebanon 2006)
But what constitutes as confidential information? There are no actual guidelines for applying what is confidential and how confidential specific documents are, and every document that contains ‘information’ as defined in the Israeli Penal Code, in part II, chapter 7, the Penal code provides a broad definition, inflicting legal sanctions on disclosing any information to an enemy where it might be useful to him (clause 111). Confidential Information is defined as any information where national security requires keeping it secret, or information relating to any matter that the government, with the consent of the parliament committee for foreign relations and security, declared as confidential. Critics to this arrangement offered an amendment, but following the Parliament’s research center’s comments, these amendments were not implemented.
The burden of proving what constitutes non-confidential information lays on the defendants in cases (see, for example, CC 1055/01 State v. Yacov), in Yacov, the court explained that while “the military censor is qualified to strike out information which is most-likely about to severely damage national security”; the penal code is wider, and applies to cases where national security requires keeping it secret.
In another interesting case, the widow of a person who worked in the nuclear research facility requested to receive the results of an epidemiological survey between the facility’s workers which the facility took. The State declined to provide the information by explaining that it relates to national security. However, when the court rejected the state claims, it expressed criticism over the state’s conduct: “the state wiggles in its arguments and cannot point to a normative authority where it draws the classification of the information. It is, according to the state, basic foundations, but these basic foundations have to be applied by the General Security Service Act, 2002, and the rules according to it (which are classified, so the state cannot disclose them to the court, but as a graceful act the state is willing to summarize them)” (CA (Tel-Aviv) 2571/01 Hanna Hizi v. State ); the court itself explained that it cannot understand classification, and the state has to acknowledge the differences between confidentiality and classification. Classification does not create basis for exclusion of evidence, and unless the state decides to exclude an evidence by means of national security according to the Evidence Act, 1971. However, in cases where the court finds the evidence may have had something to assist the party who wishes to submit the evidence, then the state shall default (OCR 2489/09 Zeev Braude v. State).
The Israeli Supreme Court deal with the question of what constitutes classified information in Vanunu (CA 172/88 Mordechai Vanunu v. State); in Vanunu, a former worker of the nuclear research facility was charged for espionage when he disclosed information regarding Israel’s nuclear activity to press agents in the UK. The supreme court decided to convict Vanunu for collecting and disseminating information to the enemy. The court analyzed this clause and explained that “He who provides information to the enemy; meaning, any information, even if it is public information arising from the press, his activities fall into clause 111”. Therefore eliminating classification need at all.
What Does Classified Information Protect? The question of what classified information protects is a difficult one to answer. Some claim that the purpose of classifying information is withholding it from foreign agents, and explain that when many people have access to certain information, it harms national security. Classifying information makes it harder for counter intelligence and foreign military forces to obtain information regarding a state’s forces, and allows it to operate where the other party does not know its rules of engagement, its powers, officers, or even defense mechanisms.
But the real question is how much this information, used by foreign intelligence, endangers national security , and does the burden of protecting this information overcome the value of keeping it secret or not.
When the classified information is the actual secret (e.g the actual location or time of a specific operation) then it is assumed (though not significant) that information about the operation that becomes available to hostile forces may lead to less successful results, at least. There are specific sets of information that are considered confidential and are not pieces of information that have (statistically insignificant) connection to current, ongoing operations or other information that if leaked may cause damage to national security.
For example, the actual existence of a specific weapon or the location where a missile fell after an air-strike cannot be considered a state secret for several reasons: the first is that it is not kept away from the public; as what the general public sees cannot be considered national secrets. For example, during the 2006 war, the military censorship requested Tapuz, Israel’s largest forum operator, to censor posts made by civilians about where Hizbullah missiles fell. Another case where information that is in the public’s plain view was considered confidential was when Parliament Member Yossi Sarid threatened that he may disclose information about weapons used by the IAF after the IAF killed and wounded dozens of Palestinians, including civilians, in weapons that were allegedly in plain view.
Another case where public plain viewed information was considered confidential was when Israel denied using phosphorous during the Cast Lead Operation of 2009, where the evidence was left in the Gaza Strip, which allowed the Goldstone committee, which inspected Israel’s activity following the operation, to find that Israel’s denial was false. So, in this case, how could the use of phosphorous be considered confidential information where there is evidence in plain view regarding the use?
Therefore, confidential information could be considered confidential as long as no public information regarding it exists. For example, the location of specific military or nuclear facilities that are located close by to cities and have road signs directing to them, could not be considered confidential information. Israeli Blogger Ido Kenan points out that Israel has a policy of withholding this confidential information in road signs presented in Arabic, and leave the confidential information only in Hebrew and English.
In conclusion, classified information in Israel is defined in an overbroad manner, containing information that may be considered in plain view and known to the general public. By acknowledging this flaw, we may understand the basis of information security and examine the weak points of such method of information security.
We believe that there has to be a difference between the classification of security mechanisms by themselves and information (data) which relates to specific, mission critical, information that is classified. The difference is between information regarding the existence and functions of a specific unit, its weapons , its history, and current plans regarding an operation.
2. Security By Obscurity, A Problem
2.1 Security By Obscurity
When trying to protect information in a digital environment, there are two popular methods used by Information Security experts. The first is Security through Obscurity: this method, which is quite similar to the Israeli Classified Information method or approach, hides all information related to security from plain view and classified it as confidential; by using this method, “a system relying on security through obscurity may have theoretical or actual security vulnerabilities, but its owners or designers believe that the flaws are not known, and that attackers are unlikely to find them”. The model bases itself on the fact that others are unaware of the activities taken and that most confidential activities could be disguised from plain view.
However, the flaws of this model are that the secrecy of the information is exactly what lets security flaws to remain secret as well. For example, GSM encryption was hacked during 2003, and again during 2009. These hacks were published to the public because they were a part of academic researches; however, in certain cases the hacker may not be so eager to publish its research. In some cases, employees or contractors may sell known exploits which were not taken care of and criminals may sell unknown exploits either to other criminals or to the company itself. Moreover, relying on a sole provider to fix the security breach could sometimes cause more problems.
The main disadvantages of Security through Obscurity may be summed up to: (1) few people inspect the system for flaws, and sometimes actually inspecting the system may be considered illegal; (2) hostile entities reviewing the security of the system do not disclose their results; (3) dependency on one vendor/provider to review and fix security breaches.
2.2 The Open Source Model.
In contrast to Security through Obscurity, Open Source advocates rely heavily on Security Through Transparency, using this method, the algorithms and software used to encrypt or protect information are known to the public, providing the public an efficient way to report security vulnerabilities, and even to propose bug-fixes. The more people have the chance to inspect the security mechanism, the safer they will be.
For example, Security firm Secunia found that more security flaws were found in the Open Sourced Firefox than in proprietary code browsers, but the number of Zero-Day unpatched flaws was significantly lower and so was the time that it took to fix any flaw. By making all of its information public, a software vendor may create better security and allow any researcher to discover flaws. Moreover, transparent security mechanisms may also deter hackers from looking how to circumvent zero day flaws in fear of being caught (See aso, David Wheeler, “Is Open Source Good for Security?”).
The Open Source Model does not ignore the basic concepts of information security, but it acknowledges their flaws and attempts to build better models.
3. Could Building a Transparent State Solve National Security?
Could we imagine a state where all public information could be deemed as non-confidential, security mechanisms would be public and open for scrutiny and confidential information would be reduced to a minimum? We believe so.
Currently, a state like Israel has to operate counter intelligence just to solve the problem of collection of plain-view information and to protect from hostile action. When operating an open source model, counter-intelligence could be abandoned and replaced with crowd sourced models, which will help to build stronger mechanisms of protection.
Moreover, removing the ambiguity relating at-least to nuclear weapons in Israel would assist deterrence and strengthen national security. Weak points in Israeli theoretical protection would be visible to the public and could be fixed quickly; moreover, the actual items that require protection could receive the needed funds and resources to protect them.
3.1 What is there to lose from revealing all classified information?
While we do not necessarily wish to reveal all information, certain information relating to means of operation and security regulations have to be declassified. For example, both the General Security Services Act and the recent Inclusion of Biometric Information and Data in Identification Documents and Database Act of 2009 state that all regulation and orders will be classified, as well as any information regarding security breaches. Moreover, when discussing the act in Parliament, security experts raised concerns over the database possible flaws, and the Minister of Interior, Eli Yishai, ordered to open the security protocols for discussion, but such discussion was never made. Keeping the database, as well as security guidelines and notifications of security breaches secret seems good in the eye of a person who thinks that an enemy may abuse such faults; however in the eyes of a security researcher, these allow zero day flaws and known vulnerabilities to be used against the database (see, for example) and allows a false feeling of security.
The only thing that may be lost when protocols, orders or regulations that remain secret are disclosed is the misconduct of an authority or its acts against the law; for example, as a result of Israel’s Freedom of Information Movement’s appeal, it was revealed that the cellular companies were required to adhere to secret regulation regarding cooperation with intelligence agencies and disclose subscriber information.
Therefore, when the governmental default approach is that there is no need for privacy unless a person has something to hide from the government (which seems to be the default approach when discussing the Israeli government, as the Biometric Database Act, the Criminal Order (Submission of Metadata) Act of 2007, and other statutes turning Israel into a surveillance state) then the default approach towards the government should be that all its secrets are meant to cover up unlawful activities.
3.2 What is there to gain from revealing all classified information?
First and foremost, the Israeli Government may regain public trust by disclosing all activities. The Israeli public, for example, strongly believes that the Biometric Database will leak, mostly due to the fact that quite a lot of sensitive data has already leaked from Government databases and that 70% of the general public does not trust database protection in Israel. A different survey by Symantec found that 60% of the people do not trust the government with their private or personal information.
The feeling of misused trust may be healed and cured when disclosing information regarding data breaches and information security to the public. But more than that, apart from public trust, the government may gain better protection of its classified information. The Israeli government may adopt what computer giants like Google and 3Com already did, and that is to pay for every security breach found.
Currently Israel has many unknown security flaws, which remain confidential until a hacker gets caught. For example, Israeli white-hat hacker Moshe Halevi (Halemo) was charged for hacking when he used a pre-paid credit card to show that the Israeli Fines and Fees Center had a bug in the URL handler that allowed resetting a person’s fines. In a detailed case (C 9497/08 State v. Moshe Halevi) Judge Avraham Tenenbaum explains why Halemo’s activity was not hacking, but was solely security checking (a similar case, CA 8333/03 State v. Mizrachi, explains that port-scanning cannot be criminal if done for a cause of security inspection). Therefore, we can argue that the state has a compelling interest to discover flaws.
3.3 The state’s approach to security flaws.
However, we see that in most cases the state prefers to withhold information from the public regarding security flaws and to litigate against persons discovering such flaws. Moreover, when flaws are found, usually adopting the Security through Obscurity approach shows that the way the state fixes the vulnerability is not only insufficient, but negligent.
In one case, white-hat hacker Halemo discovered that the Israeli Court System’s website discloses Judge’s ID Numbers (equivalent to Social Security numbers). The way it disclosed them was that the URL Source of the Judge’s page in the website was his ID number. After the flaw was exposed, the state went to fix the flaw, and replaced the ID with a Base-64 representation of the number.
However, if we require the state to disclose its means of security it would have to disclose how the judges ID numbers were encrypted or protected, and therefore every person would have understood that neither plain-text nor base-64 are good enough mechanisms to protect sensitive information.
4. Applying Software Solutions to State Secrets: A Conclusion.
We believe that not all information has to be public. There are things that are better off secret. However, if we learn from information security methods, we must acknowledge that better security could be achieved when disclosing more information to the public. Applying the open source model of information security allows transparency in decision-making, better algorithms, less resources on counter-intelligence and more resources to allocate to what is mission critical information.
Moreover, better trust could be gained between governments and citizens, reinforcing the social contract and allowing better results in political participation.
Currently, governments over trust security through obscurity when operating mission critical processes, and therefore, when flawed, the flaws and results are enormous. Utilizing open source models could prevent mishaps such as Israel’s phosphorous use, George Bush’s Weapons of Mass Destruction lie and Israel’s racial profiling in Airports as a mean of security.
Israeli racial profiling is such a great example, as it is highly efficient nowadays and even better than the US TSA guidelines but bases itself mostly on the assumption that Jewish nationals may not be considered a threat to national security but Arabs may (HCJ 4797/07 The Israeli Association of Civil Rights v. The Terminal Security Authority, Pending decision). As long as the security guidelines were secret, it seemed amazing that no security flaw occurred. However, now, that the guidelines are known and understood, it is easier to design a mechanism to circumvent them. Therefore, even adopting new guidelines will be useless, as they are inefficient (unless based, again, on racial profiling).
Therefore, in order to regain national security, Israel will have to change its approach to the Open Source Model before a major security event occurs that will make it understand that this is the only option. Staying in a Security through Obscurity approach could protect confidential information, but it cannot protect national security.
