True Names and the Opening of the Cyberspace Frontier (6 page)

The actual details involve some complicated math and need careful thought to get straight, which this article cannot cover. Bruce Schneier's
Applied Cryptography
has a good explanation of how Chaumian digital cash works, and
Scientific American
has also carried some good articles.

It is often claimed that “digital currencies” will not gain widespread acceptance, let alone the support of governments. If digital money is viewed as a transfer mechanism, and not as a competitor to currency or specie (gold, silver, etc.), then the support of governments is less of an issue, perhaps even a non-issue, because banks have done quite well without explicit governmental sanction of their instruments. And in the international realm, there already is not much of a governmental role: banks have worked out mechanisms for dealing with each other, and for dealing with entities with a reputation for misbehavior. As we will see, international trade represents a kind of anarchy.

There are many reasons for using untraceable digital cash. Some people simply prefer to pay cash for various reasons, and see no reason why electronic transactions should have more traceability than ordinary folding-money transactions have. Others fear the compilation of dossiers on spending habits, travel agendas, and so forth. Untraceable digital money protects the privacy of economic transactions, just as cash does today. With increasingly powerful networks of ATM and check-processing systems, the development of “shopping profiles” is a concern for anyone interested in privacy. Having insurance companies and employers gaining access to purchasing habits is undesirable; such access could, at its extreme, lead to law enforcement midnight raids on persons suspected of various crimes because of legal purchases they might have made. Untraceable digital money provides protection against this.

Making automated toll-road payments with untraceable digital cash is one obvious use. Digicash is working with European governments to deploy digital money for this sort of application.

There are, of course, various transactions involving anonymity, digital pseudonyms, and illegal items that only an untraceable digital cash system makes possible. And some novel applications are new. For example, “perpetual trusts” could be constructed by purchasing a large number of digital money instruments, perhaps being converted regularly to other such instruments. Because they are untraceable, there is no means of, say, canceling the numbers to stop the perpetual trust. Thus, as a hypothetical, no one—certainly not the bankers—will know that which of the instruments are part of the perpetual trust Bill Gates creates in 2010 with ten billion dollars … and this trust could still exist a century later, untouched by taxation and not even really domiciled in any particular nation. Contracts using such digital money instruments could similarly be of this “fire and forget” sort. Thus can fortunes be directed toward specific purposes, beyond the reach of governments. (For the curious, digital time-stamping and cryptographic timed-release techniques are needed to insure that the humans involved don't violate the contract originally set up.)

There are, of course, many reasons
not
to use untraceable digital cash. Businesses typically need to show records of expenses to deduct against gross sales. The simplest example of this involves anonymous payments to employees: few corporations would be interested in doing this, even if they satisfied themselves that they wouldn't get caught, because they then could not use the employee expenses as a deduction against raw income. (One can imagine many situations where an employer
would
be interested in such arrangements, and under-the-table payments are common practice in certain types of businesses.)

There is still the possibility of fraud, of dissatisfaction with transactions, and of improperly completed transactions. Cryptography obviously cannot completely eliminate such disputes. But various measures, such as reputation-rating services, digital signatures, etc., should work fairly well in controlling these kinds of problems. Trade has been conducted for millennia without governments playing a central role; in fact, international trade is often cited as an example of anarchy in action, as clearly the laws of any one country are not easily applicable. That trade works so well is evidence that actions have consequences, that repeat business matters, and that even in a relative anarchy, behavior matters. An excellent survey of this kind of trade anarchy is contained in Bruce Benson's
The Enterprise of Law.

The argument often made by critics of untraceable e-cash, that issuers will renege or abscond, refusing to honor their instruments, ignores the nature of e-cash. Because e-cash is untraceable, an issuer never really knows when he's merely being “tested” by a rating service (or, more direly, when the client might be a member of the Mafia!). Reliability testing and reputation ratings are important.

True digital cash—the fully untraceable form—admittedly will allow some new channels for criminal activity. Privacy has its price. The ability of people to plot crimes and commit them behind closed doors is obvious, and yet we don't demand secret cameras in homes, apartments, and hotel rooms. Some of the disadvantages of anonymous systems will be discussed later, along with some of the proposals by various governments to limit or even completely ban strong cryptography.

The Surveillance Society

Imagine you are entering a bar or nightclub, or a movie. You are asked to produce identification as proof that you are of legal age. Currently, these “credentials” are presumably only glanced at briefly. With the advent of computer scanners, bar codes, and networks, the very real possibility exists that such credentials will be scanned, read, and fed into various databases. Maybe for customer profiling, maybe for compliance auditing, maybe for other reasons. But the effect is that one's movements, habits, and preferences are now in a database, perhaps even fed to the local police (as is the custom in many countries). Even if the collected data is not explicitly planned for a dossier, or for the government, a trail is still created, and this presents serious problems, especially as networks and computers get much faster.

David Chaum, along with his other work, has also developed schemes for presenting a credential of some sort without revealing identity. Though this sounds impossible, modern cryptography provides an approach. Think of it as a sealed envelope with a movable transparent window that can be moved over, say, an “age” field. The owner of such a credential could present proof that he is of some age, or past some age, without providing his identity or any other information. How this works, and how forgeries are prevented, is beyond the scope of this chapter. Cryptographic protocols are used, and biometric authentication is generally needed, to prevent such a credential from being easily lent or sold to others.

One obvious use is for automated toll-road tokens that can be read remotely, either authorizing the holder to travel on the road, or, using digital cash, make a payment remotely. The dangers of having one's movements on toll roads compiled into records is obvious to nearly everyone, though Singapore has adopted just such a citizen-unit-tracking system!

This is a good example of how technology can provide the kind of protection that well-meaning “privacy laws” cannot actually provide. While special interest groups lobby the government for new laws and new wrinkles on old laws, technology can directly provide the protection many want. For example, which approach better solves the problem of people using scanners to monitor cellular telephone conversations: passing more laws saying such monitoring is illegal (except for the police), or adding encryption to cell phones? A basic credo of the Cypherpunks movement has been that technological solutions are preferable to administrative or legislative solutions.

The growing use of government-approved picture IDs for travel is becoming the modern equivalent of travel documents in the U.S. While I cannot see a situation in which citizen-units are ever told they may not travel without authorization, I can quite easily see the situation emerging in which airlines, bus companies, car rental agencies, hotels, and gas stations are expected to “run your card through.” This is already the case with many hotels and nearly all car and truck rental agencies demanding credit cards (partly to insure payment, but also for law-enforcement purposes). This produces a de facto movement-tracking system. Expect more scrutiny, perhaps even time-consuming and hassling scrutiny, for those who try to pay in cash and for those who are reluctant to have their ID cards run through the system. Since 1995, airlines have insisted on picture IDs, on orders of the government.

As with the government interest in true names and the naming process for tracking, such ID cards are an essential tool for tracking movements, collecting taxes, and establishing dossiers on citizen-units. Credentials without identity are an important technology to have and to deploy widely. A recurring theme here is that technology, not so-called privacy laws (from which governments nearly always exempt themselves anyway), is the best protection against such a surveillance state.

Data Havens and Information Markets

Another science fiction writer, Bruce Sterling, popularized “data havens” in his 1988 novel
Islands in the Net.
He focused on
physical
data havens, but cyberspace data havens are more interesting, and are likely to be more important. That they are distributed in many legal jurisdictions, and may not even be traceable to any particular jurisdiction, is crucial. A data haven is a place, physical or virtual, where information may be stored or accessed. The usual connotation is that the data are illegal in some jurisdictions, but not in the haven.

Data havens and information markets are already springing up, using the methods described to make information retrievable anonymously and untraceably. Using networks of remailers and, of course, encryption, messages may be posted in public forums like the Usenet, and read by anyone in the world with access, sort of like a cyberspatial “Democracy Wall” where controversial messages may be posted. These “message pools” are the main way cyberspatial data havens are implemented. Offers may be in plaintext, so as to be readable by humans, with instructions on how to reply (and with a public key to be used). This allows fully untraceable markets to develop.

It is likely that services will soon arise which archive articles for fees, to ensure that a URL (Uniform Resource Locator) is “persistent” over a period of many years. Ross Anderson's “Eternity Service” provides a means of distributing the publication of something so that even later attempts to withdraw all copies are thwarted. This has obvious value in fighting censorship, but will also have implications when other types of publication occur (for example, a pirated work would not be withdrawable from the system, leaving it permanently liberated).

Examples of likely data haven markets are credit databases, doctor and lawyer databases, and other heavily regulated (or even unallowed) databases: information on explosives, drug cultivation and processing, methods for suicide, and other such contraband information. Data havens may also carry copyrighted material, sans payment to holders, and various national and trade secrets.

As one example, the “Fair Credit Reporting Act” in the U.S. limits the length of time credit records may be kept (to seven or eight years) and places various restrictions on what data may be collected or reported. What if Alice “remembers” that Bob, applying for credit from her, declared bankruptcy ten years earlier, and ran out on various debts? Should she be banned from taking this into account? What if she accesses a database that is
not
bound by the FCRA, perhaps one in a data haven accessible over the Net? Can Alice “sell” her remembrances to others? (Apparently not, unless she agrees to the various terms of the FCRA. So much for her First Amendment rights.) This is the kind of data haven application I expect will develop over the next several years. It could be in a jurisdiction that ignores such things as the FCRA, such as a Caribbean island nation, or it could be in cyberspace, using various cryptographic protocols, Web proxies, and remailers for access.

Imagine the market for access to databases on “bad doctors” and “rip-off lawyers.” There are many interesting issues involved in such databases: inaccurate information, responses by those charged, the basis for making judgments, etc. Some will make malicious or false charges. This is ostensibly why such databases are banned, or heavily regulated. Governments reserve the right to make such data available. Of course, these are the same governments that falsify credit records for government agents and that give the professional guilds like the American Medical Association and the American Bar Association the power to stop competitors from entering their markets.

Information markets match potential buyers and sellers of information. One experimental “information market” is BlackNet, a system I devised in 1993 as an example of what could be done, as an exercise in guerrilla ontology. It allowed fully anonymous, two-way exchanges of information of all sorts. The basic idea was to use a “message pool,” a publicly readable place for messages. By using chains of remailers, messages could be untraceably and anonymously deposited in such pools, and then read anonymously by others (because the message pool was broadcast widely,
à la
Usenet). By including public keys for later communications, two-way unreadable (to others) communication could be established, all within the message pool. Such an information market also acts as a distributed data haven.

As Paul Leyland succinctly described the experiment:

Tim May showed how mutually anonymous secure information trading could be implemented with a public forum such as Usenet and with public key cryptography. Each information purchaser wishing to take part posts a sales pitch and a public key to Usenet. Information to be traded would then have a public key appended so that a reply can be posted and the whole encrypted in the public key of the other party. For anonymity, the keys should contain no information that links it to an identifiable person. May posted a 1024-bit PGP key supposedly belonging to “Blacknet”. As May's purpose was only educational, he soon admitted authorship.