On Understanding Biometrics

A decade or so ago, I had an extended dialogue with David Clark, of the Clark-Wilson Model, about biometrics. He knew that the remedy for a compromised password was to change it. Since he knew that biometrics could not be changed, he could only understand how they worked for about fifteen minutes at a time, about the same length time that I can understand the derivation of an RSA key-pair.

Unlike Passwords, biometrics do not rely upon secrecy, they do not have to be changed simply because they are disclosed. Biometrics work because, and only to the extent that, they are difficult to counterfeit.

We have all heard about the attacker who picked up a latent fingerprint in gelatin and replayed it, or the image scanner that was fooled by a photo. Good biometric systems must be engineered to resist such attacks.  For example, Google has patented a (challenge-response) scheme to resist replay in a facial recognition system by directing the user to assume one of several expressions that the system has been trained to recognize.  

While the fundamental vulnerability of passwords are replay and disclosure, the fundamental vulnerabilities of biometrics are replay and counterfeiting. These vulnerabilities are limitations on the effectiveness of the mechanism, rather than fatal flaws. What we must ask of any biometric system is how does it resist replay and counterfeiting.

At a recent NY Infragard meeting there was discussion of biometrics that illustrated that this confusion persists. In this case, at least one person seemed to be convinced that the secrecy of the stored reference had to be maintained in order to preserve the integrity of the system.

As with properly stored passwords, one cannot go from the stored reference to what one must enter. We solved the problem of disclosure of the password reference by encrypting it with a one-way transform at password choice time. (in practice, all too many passwords are stored in a reversible form.) At verification time we apply the same transform to the offered password and compare the encrypted versions. By definition of "one-way transform," it is not possible to go from the stored reference to the clear password.

We do not store an image of the face, the password, or a recording of the voice. Instead we store a one-way, but reproducible, transform. To make sure that the transform is reproducible, we may collect multiple samples so that we can test for reproducibility.

However, as with passwords, biometrics are vulnerable to replay attacks. While we cannot discover the biometric from the reference, we might capture an offered instance and replay it over and over.

Like passwords, biometrics may be vulnerable to brute force attacks, but unlike passwords, they are not vulnerable to exhaustive attack, if only because it is impossible to try every possibility. While a password reference can be stored in 8 or 16 bytes, a biometric reference may be hundreds or low thousands of bytes.  In an exhaustive attack against a password, each unsuccessful trial reduces the uncertainty about the correct answer, in part because we know the max size. This may not be true about a brute force attack against a biometric where the maximum size may be arbitrary.

For most purposes we use brute force and exhaustive as though they were synonymous but they really are not. In brute force, we submit a sufficient number or trials to succeed in finding a (false) positive. An exhaustive attack is a special case of a brute force attack in which we are trying to find one integer in a known set. While the reference for a biometric may be too large to be exhausted but there are many values that will fit.

This introduces the issue of false positives. There is only one password that will satisfy the transform, it is at least one integer value away from those that do not satisfy. We key in the integer. However, we "sample" a biometric; there will be many biometric samples that will fit. Depending upon the precision of our system, it might even be possible to dupe the system, a false positive. On the other hand, it is also possible for a given sample of a valid biometric to be rejected, a false negative.

Biometric systems can be tuned so that they achieve an arbitrary level of security; we are looking for a transform that minimizes both false positives and false negatives. Unfortunately we reduce one at the expense of increasing the other. That is to say, the less likely it is for the system to permit a false positive, the more likely it is to generate a false reject. We tune the mechanism to achieve an acceptable ratio of on to the other for a particular application and environment.

My preferred biometrics are the visage, i.e., the face, and the voice. These share the advantage that they can be reconciled both by a computer and by non-expert human beings. Infants can recognize their parents, by face and voice, by the age of six months; it has survival value. Many share the experience of recognizing someone, that we have not seen in years, from one or two words, spoken over the telephone.

Until very recently, machines could not reconcile faces as fast humans, indeed fast enough for many applications. However, Google now has software that can not only authenticate an individual from an arbitrary image but identify them within seconds.

For most of the time that they have been in use, fingerprints could only be reconciled by an "expert," but we now have computers that can do it even better than the experts. In fact, recent studies using these computers have suggested that even these experts are all too fallible. Nonetheless, non-experts can independently verify fingerprint identification.

Think about DNA. While it discriminates well, it contains so much information that it takes a long time to reconcile and the results cannot be independently verified by amateurs. To some extent we will always be dependent upon instruments and experts.

Since biometrics share a vulnerability with passwords to replay, a password plus a biometric does not qualify as "strong authentication." Therefore, the preferred role of biometrics is either as an identifier or as an additional form of evidence in a system of strong authentication, one in which another mechanism, e.g., a smart token, is used to resist replay.
Because there is only a vanishingly small chance that two samples of a biometric will be identical, any sample that matches one previously submitted could be thrown out as a possible replay.

Part of the special knowledge that identifies us as security professionals, and for which we are paid the big bucks, is that knowledge about the use, strengths, and limitations of biometrics.

Who is Responsible for Security?

On September 30, 2011, SANS Institute NewsBites reported the following story:

--European Union to Introduce Liability Rules for Cloud Vendors (September 28 & 29, 2011) The European Union (EU) plans to introduce the "Binding Safe Processor Rules," which would hold vendors of cloud services in the EU liable for data security breaches. Vendors would sign up for what amounts to an accreditation. Consumers are likely to feel safer doing business with a company that is willing to stand behind its services. The rules are an update to the Data Protection Directive. The companies will be required to demonstrate their compliance with certain data protection standards for approval under the rules. Current law holds data owners responsible for data loss.

They cited two sources, SC Magazine and V3.co.UK, The Frontline.

http://www.scmagazine.com.au/News/275173,eu-cloud-vendors-liable-for-breaches.aspx

http://www.v3.co.uk/v3-uk/the-frontline-blog/2112906/eu-rules-allow-cloud-companies-legally-customer

The NewsBites editor added the following comment by me.

[Editor's Note (Murray): The devil is in the details and the rules may be helpful. However, the idea that one can transfer the responsibility for protecting the data from the owner to the custodian by fiat, or any other way, is absurd on its face. The decisions about protecting the data cannot be separated from the decisions about collecting it and using it.]

While I confess I misread the details, they are where the devil is hiding. It turns out that this rule is nothing like it sounds either in its name or in this report. Instead it was sought by Amazon, Google, and others to say that EU enterprises may rely for security of their data upon service providers that are certified by an EU country as complying with these rules and without regard to location. It is a response, in part, to the fact that Europeans will not do business with US service providers because they are subject to the USA Patriot Act. They are concerned that they would be accused of improper reliance. The EU has never been happy with the idea of data on Europeans being stored in the US.

This week NewsBites reported this story:

--Stanford Hospital Pins Breach Responsibility on Third-Party Billing Contractor (October 6, 2011) Stanford Hospital & Clinics says that a data security breach that compromised the personal information of 20,000 patients is the fault of a third-party contractor. One of the patients filed a US $20 million lawsuit against Stanford following the breach disclosure last month. The data were exposed because a spreadsheet handled by a billing contractor somehow was posted to a student homework help website. The compromised information includes names, diagnosis codes and admission and discharge dates.

http://www.computerworld.com/s/article/9220626/Stanford_Hospital_blames_contractor_for_data_breach?taxonomyId=17

Now, I have to tell you, the Hospital tells a really great tale. Mind you, it does not excuse them for the breach. However it might have confused a jury if they had not attempted it to try it out in the media first.

Seems they turned the data over to a collection service, MCSC, in encrypted form. This is allowed under HIPAA rules but requires that they have security of the data as part of their agreement with the service provider.

Needless to say the collection agency, MCSC, decrypted the data. It converted it to a spread-sheet before turning it over to an "unauthorized" third party, This third party posted it, as an attachment to a request for assistance, to a site called Student of Fortune where it remained for a year. Student of Fortune is a site where students can solicit assistance with their homework assignments. It seems this third party wanted assistance with a graphical representation of the data in the spreadsheet. It would probably be unfair for one to infer that someone familiar with such a site is a recent student. There must be some truth here. You can't make this stuff up.

It seems clear that there is plenty of blame to go around here. However, the question is not blame but responsibility, ethical, legal, financial, and otherwise.

Public and private enterprises are increasingly relying upon contractors and other enterprises, "partners," to carry out duties and responsibilities that historically have been performed by employees and within the enterprise. Therefore, it is timely to revisit the question of responsibility.

Both of these stories suggest that the responsibility rests with the custodians of the data, The first story suggests that the responsibility can be assigned to the custodian by order of the state or the consent of the custodian. The second suggests that the responsibility moves with the data.

Ultimately, the legal questions raised by these stories will be decided by courts. I can hardly wait. I am a great fan of court records and decisions. While subject to error, they are much more reliable than the statements of the parties.

In Information Assurance, we have traditionally assigned protection duties and responsibilities in terms of roles, i.e., management, staff, owners, custodians, and users. We have argued that, by definition and default, the responsibility to protect the data rests with the "owner," the manager responsible for all the decisions about the data.

For example, the owner makes the decision to collect and store the data. The owner, again by definition, makes the decisions about who can use the data. The owner makes the decision as to the sensitivity of the data, how much to spend on protection and how much risk to accept. The owner's responsibility includes communicating these decisions to custodians and users.

It is difficult to see how this control and discretion can be separated from the responsibility for its exercise.

Our colleague, Bob Johnston, likes to argue that "When entrusted to process, you are obligated to safeguard." However, as a custodian I would respond by asking how much and at whose expense? Clearly a custodian would not want to spend more than the owner would and would expect to be reimbursed or compensated for what he does spend.

What is really at issue here is how we identify and select custodians, describe their duties, compensate them for those duties, what penalties they must pay for breach of those duties, and to whom. Obviously, this begins with negotiations between the owner and the custodian. I will continue to argue, both as matters of definition and practicality, that the responsibility for the results, the success, of those negotiations must start and end with the owner.

As a matter of law and good public policy, we want the responsibility in the same hands as the discretion. The alternative would permit the owner to pick the low cost service provider and then escape responsibility for any consequences. One might call that moral hazard.

Service providers are in the role of custodians of the data. Their duty is to the owner of the data, the party that pays them, not to the subjects of the data. They must be diligent in the execution of the duties that they have agreed to and for which, in part, they are being paid.

Stanford Hospital had a duty to their patients to protect the data. That duty did not go down when, for their own convenience and efficiency, they decided to give a copy to another party, a party of their choice. That they encrypted it for purpose of transfer, did not protect it from that agency, to whom they also gave the key. The agency's duty was to Stanford Health, to protect the data in accordance with their agreement, the provisions of which we are left to guess. While it is unlikely that Stanford Hospital specifically contemplated the possibility that MCSC would give a copy to a contractor, their agreement should have resisted it.

One might argue that as a collection agency, the agency owed a duty to the subjects of the data. However, it would be hard to argue that that duty relieved Stanford Health of its responsibility..

As security staff, some for the owners, some for the custodians, our role is to assist the business managers and lawyers in expressing the security requirements in such a way that all parties understand their duties and are likely to discharge them in manner that will produce the intended results. Our job does not stop there; we must go on to measure and report the results, note variances from the expected and intended, and recommend timely corrective action on a timely basis. "Timely" is before, rather than after, any breach. To the extent that this is difficult, we are called professionals and are paid the big bucks.