Anonymity on the Internet and Ethical Accountability

Terrell Ward Bynum

[This brief position paper summarizes a presentation by the author at the Conference on Anonymity on the Internet organized by the American Association for the Advancement of Science in autumn 1997 (funded by the National Science Foundation).]

In 1994 Helen Nissenbaum published an article in Communications of the ACM entitled “Computing and Accountability.” In that article Nissenbaum spelled out important relationships among the concepts of responsibility, blame, and accountability. She also made a strong case for the view that accountability “encourages diligent, responsible practices” and provides “the foundation for just punishment as well as compensation for victims.” Nissenbaum noted:

Responsibility and blameworthiness are only a part of what is covered when we apply the robust and intuitive notion of accountability.... When we say someone is accountable for a harm, we may also mean that he or she is liable to punishment (e.g., must pay a fine, be censured by a professional organization, go to jail), or is liable to compensate a victim (usually by paying damages). In most actual cases these different strands of responsibility, censure, and compensation converge because those who are to blame for harms are usually those who must “pay” in some way or other for them.

In that same article, Nissenbaum identified four “barriers to accountability” associated with current computing practices. These include (1) the problem of “many hands” in which a wide variety of individuals and institutions can be involved in the design and creation of a computer system, (2) the ease with which people accept so-called “bugs” in a computer program as unavoidable, (3) the tendency to treat computers as “scapegoats” for a variety of errors, and (4) the desire to own a computer program without accepting liability for it.

I would like to suggest that yet another “barrier to accountability” exists; namely, anonymity on the Internet. If a person gets onto the Net and anonymously engages in some kind of harmful activity (e.g., malicious hacking; defamation of character; extensive “spamming”; dissemination of computer viruses; industrial espionage, etc.), it may not be possible to hold such a person accountable. We therefore would be unable to assess blame, enforce the law, prevent repetitions, secure compensation, or gain other benefits of accountability.

So, given the many social and ethical benefits of accountability, one might be tempted simply to argue that anonymity on the Internet should be banned – that the identity of anyone on the Net should always be immediately available wherever he or she goes in cyberspace.

This extreme view, though, seriously conflicts with privacy; and there are many circumstances in which one wants privacy to be preserved – for example, engaging in discussions of “sensitive” topics such as, HIV, abortion, gay lifestyles, breast cancer, politically unpopular topics, etc. Even when we are simply “surfing” the Web – browsing topics of interest, shopping in cyberspace, reading the news, doing research, job hunting, whatever – we want to preserve our privacy. For if we could be tracked wherever we go in cyberspace, and a record were to be kept of how long we stayed at each site, what topics and organizations we spent time with, what we purchased, how much money we spent, and so forth and so on, a revealing and privacy-invading profile could be created on each of us.

It is clear, then, that there are circumstances in which it would be socially beneficial to preserve anonymity on the Internet, as well as circumstances in which anonymity on the Net would be harmful and undesirable. How do we know which case is which? And, more importantly, how do we secure privacy and anonymity when they are desirable, while nevertheless preserving accountability when that is needed? There is an urgent need to answer these questions.

Personally, I would start my search for answers by assuming that privacy is the default – that privacy is to be preserved wherever possible and balanced against other human values like justice and security. And clearly, to the extent that anonymity is necessary to privacy, it should be preserved in those cases where privacy should take precedence.

In our search for answers, one helpful way to deal with circumstances in cyberspace is to argue by analogy using similar circumstances that are not in cyberspace. This method is often very productive because aspects of “ordinary” circumstances frequently carry over into cyberspace with similar results. For example, in non-cyberspace contexts, we regularly sacrifice some of our privacy by exchanging it for confidentiality. Thus, in order to get the benefits of medical doctors, psychologists, accountants, religious counselors, etc., we share private information with them on condition that they keep it confidential. Although, strictly speaking, privacy is lost in such a case, many of the benefits of privacy are preserved through confidentiality. In addition, in a case like this, if it suddenly becomes necessary to hold someone accountable, the confidential information can be disclosed in ethically defensible ways to appropriate people.

Similarly, the idea of having “trusted third parties” in cyberspace – agents to whom one entrusts private information on condition that it be held in confidence – can be used to gain the benefits of privacy and anonymity in cyberspace, while at the same time preserving the possibility of holding people accountable in extraordinary circumstances. For example, when an individual wishes to make a business transaction in cyberspace, he or she could do so through a trusted third party, who pledges not to reveal to others any information about the customer or the transaction. In extraordinary circumstances, though, confidential information could be revealed to the proper persons in an ethically defensible manner. Perhaps many different activities on the Net could be “confidentialized” through a trusted third party – or through various kinds of clever software – thereby preserving the benefits of anonymity, while retaining accountability under extraordinary circumstances.

In the past, society has developed many ways to permit anonymity under normal conditions, but at the same time make it possible to “unanonymize” a person in extraordinary circumstances. Thus, in non-cyberspace contexts there are many times when a person is effectively anonymous – walking down a crowded street, shopping in a store and paying with cash, going to the movies, fishing in a stream, etc. Just as society has developed various ways to identify “anonymous” agents when the need arises (e.g., eye witnesses, finger prints, opening sealed records, etc.), perhaps it will be possible to devise reliable ways to “unanonymize” someone in cyberspace who normally could go on his way unnoticed. What are the cyberspace equivalents of eye witnesses, finger prints, unsealed records, etc.?

Back to the top

Home > Research Resources > On the Emerging Global Information Ethics > Anonymity on the Internet

The Research Center on Computing & Society
at Southern Connecticut State University
501 Crescent Street • New Haven, CT 06515
Director: (203) 392-6790 • e-mail: webmaster@computerethics.org

© 2000 – 2007 – Research Center on Computing & Society