Welcome to the year of living less dangerously

In the third of a series of articles on Internet law, Andrew Charlesworth looks at libel online. This could be the year in which defamation law in the England and Wales begins to catch up with electronic media, making it easier for Internet operators to know where they stand.

Defamation law can be divided into two strands. The law of slander concerns defamation by speech. The law of libel concerns defamation in writing. It seems that libel law will be applied to the various forms of Internet usage, including the creation and management of web sites.

Libel, say the lawbooks, "consists of a defamatory statement or representation in permanent form". Anything temporary and audible only is slander. Defamatory statements in books, articles, newspapers and letters are libels. In the early days of computer law, it might have been possible to argue before the courts that electronic communications or publishing were not permanent and thus could not be libellous. But the courts have become more computer literate, and it is easy to point out that the products of electronic communications or publishing are often printed for dissemination and reading, so it is unlikely that this line of argument would succeed.

The allegation complained of must be defamatory as opposed to vituperative or abusive. Those involved in flame wars on the Internet are frequently abusive about their opposite numbers. But merely venting one's feelings, even if they injure the other person's self esteem, does not constitute libel. Defamation is only made out where, as the result of the publication of an untruthful or inaccurate statement, the plaintiff is held in lower esteem by others. To describe someone as a tiny-minded cretin in an email or on a Web page is unlikely to attract a libel suit. To describe the same person as an adulterer or a child molester, without convincing evidence, almost certainly will.

The defamatory statement must recognisably refer to the plaintiff. That is, the ordinary person in the process of scanning the newspaper, email, bulletin board or web page would upon reading the statement be led to believe that the plaintiff was being referred to. The plaintiff need not be named. It is enough that his or her identity be readily ascertainable from the statement.

The defamatory statement must have been made known to others - it must have been published. Communication to the defamed party is not publication, as libel is concerned with how third parties view the plaintiff. It has been said that "the question of publication of a libel contained in a letter will depend on the state of the defendant's knowledge, either proved or inferred, of the conditions likely to prevail in the place to which the libel is destined". In the case of an electronic mailing list the individual posting the communication will be well aware of the fact that it will be widely disseminated by the listserver. An individual posting material on a bulletin board system will similarly be aware that it will be available to anyone who dials up that BBS. Web pages might well have the widest potential dissemination of all. The law has been that "every repetition is a fresh publication giving rise to a fresh cause of action against each successive publisher". The editor, printer and publisher of an article and even mechanical distributors such as bookstalls may be liable. This is currently subject to the defence of unintentional defamation. It might be argued that BBS sysops are liable for postings on their boards, and that a moderator of a moderated email list (where the messages are read by the moderator before posting to those on the list) almost certainly would be. Similarly the owner of a Web server would appear to be caught by this, and even possibly an individual providing a link to the Web page in question.

A draft Bill, sponsored by the Lord Chancellor's Department, aims to clarify some of the issues. It may lead to the enactment of new legislation during 1996. The department has been encouraged to act by the fear that a large number of net related cases will soon come to court, and that existing defamation law will be unable to cope. Reactions to the original consultation document were mixed. It was felt that the provisions it envisaged would in fact discourage online service providers such as net access providers, proprietary hosts and bulletin board operators from monitoring the material that they carry, even where they do so at present.

Under the present legal regime it is not clear whether such an Internet service provider could be held responsible for the nature of the material transmitted via its services. This uncertainty arises out of confusion over the status of service providers. If they are held to be publishers in the traditional sense, they can be held responsible for material they distribute. However, if they are held to be similar to operators of a basic telecommunications service or a postal service, such bodies are normally classed as "common carriers" and enjoy greater protection against defamation actions. No UK service providers have been sued for defamation, so the issue remains undecided.

The draft Bill (available from HMSO, price Pounds 3.75) creates a new statutory defence which will be available to distributors, printers and others who do not have primary responsibility for a defamatory publication, provided they exercised reasonable care in relation to that publication, and they neither knew nor had reason to believe that their acts contributed to the publication of defamatory material (S1).

This defence is not available to authors, editors or publishers. S1(3)(c) and S1(3)(e) of the draft Bill exclude from these categories those people who merely process, make copies of, distribute or sell any electronic medium in or on which a defamatory statement is recorded; operate the equipment by which it is retrieved, copied or distributed; or operate a communications system by which it is transmitted or made available. The latter exemption, however, requires that there be no effective means of control by the operator over the person making the statement.

The apparent problem with these exemptions lies in section five of the Bill which provides grounds on which to determine whether a person has taken reasonable care or has reason to believe that his actions may have caused or contributed to the publication of defamatory statements. The three grounds are:

* the extent of his responsibility for the content of the statement or the decision to publish it

* the nature or circumstances of the publication, and

* the previous conduct or character of the author, editor or publisher.

Suppose a service provider decides to monitor messages on a BBS or pages on a Web server, and where it feels it necessary, to censor them. It is hard to tell whether it can use the new defence of not being responsible for publication, should a defamatory message slip through. It will, after all, have moved from being a passive carrier of information to performing some sort of editorial function. A Web site which monitors student Web pages might also find itself liable for defamatory statements on unmonitored staff pages. If the Bill passes into law in its current form, it will be interesting to see how the courts handle this.

Given the amount of information which may pass across a bulletin board, or be stored and adapted on a web server, attempts to perform 100 per cent effective scrutiny for defamatory statements are unlikely to succeed. It seems likely that the best approach for service providers, if the Bill becomes law, would be to abdicate any responsibility for content. By refusing to monitor content, they would reduce the risk of being successfully sued.

But it would appear from a reading of S1(5) of the Bill and its explanatory memorandum that a service provider who was aware that someone had previously published defamatory material might be running some risk of liability if it did not check that person's subsequent output for potentially defamatory content. This may pose a dilemma for both net service providers and institutional Web sites.

This issue has already been litigated in the United States, where service providers who performed a monitoring or censorship function have, as a result of that activity, been found liable for defamation. This happened to the online service Prodigy. By contrast those who have not performed such a function have in effect been held to be passive carriers and not liable for content. However in the US, the absence of any form of "all reasonable care" test allows a service provider the unfettered option to not monitor traffic, and thus effectively to be classed as a common carrier. Forthcoming legislative changes notwithstanding, it seems likely that the courts in England and Wales would be inclined to accept that it is possible to libel individuals on BBS, mailing lists or Web pages by posting untruthful and damaging statements about them. A person posting untruthful and damaging statements about another on a list, BBS or Web page where such a message would be read by others in the country, and where this would be damaging to the plaintiff's good name or reputation, would almost certainly be liable to a libel action.

The issue of a Briton libelling a US citizen is perhaps less clear as this might be considered by the courts to be out of their jurisdiction. However in the recent case Godfrey v. Hallam-Baker (unreported) where the claim was for damages for libel or alternatively slander regarding seven Usenet messages posted in 1993, the defendant apparently worked at Cern in Switzerland. This case appears to have been settled out of court in June 1995 when it was reported that Laurence Godfrey accepted undisclosed damages from Philip Hallam-Baker.

Publication on the net will involve publication in a number of different countries. This may encourage jurisdiction hopping. As the law of defamation differs in each country, it will be advantageous for the plaintiff to choose the country with the most punitive libel laws and the fewest defences. A plaintiff libelled on a Web page accessible to the UK and US education domains is likely to sue in the UK, where the standard for defamation is easier to meet with fewer defences.

The nature of the electronic dissemination itself might be a factor. Suppose an individual sends a libellous communication to a listserver in Canada, intending all the subscribers of the list (including some in the UK) to receive the message, as opposed to posting the libellous communication on a BBS in the UK, where someone in Australia dials up and reads it. In the first case the individual is deliberately placing the material into a foreign jurisdiction. In the second that intent would appear to be lacking. The nature of the Web, however, would appear to place all such defamatory material placed on Web servers into the first category.

The law of defamation poses an interesting problem for the concept of free speech. Setting up a bulletin board or Web server is not particularly expensive. However, the cost of defending a legal action can be high. Concern has been expressed that the law of libel may be used to stifle what many net users have long considered to be their right to absolute free speech on the networks, and that this will in turn have a chilling effect upon the willingness of individuals to carry out certain forms of electronic publishing. It could also make academic institutions less willing to allow staff and students a free hand in this area.

Given the choice between suing a student and suing a higher education institution for libel, a plaintiff is almost always going to go after the institution. For Web site operators the most stringent approach to this problem is simply to bar all but official institutional Web pages. However, as most institutions allow their students access to email, and all the cases of electronic libel so far have concerned email, this would seem to be short sighted. All members of an institution could create Web pages, following guidelines. Personal Web pages should not be allowed to pass themselves off as representative of the institution. Use of official logos or designations should not be permitted. The institution should disclaim editorial responsibility for personal Web pages, but have a clear procedure for removing material brought to its attention as potentially libellous, until legal advice can be taken as to its status. This should help to provide a shield against liability, should anyone use the institution's machines to disseminate defamatory material. However, this will always depend on the response of the institution once the defamation is brought to the attention of the authorities.

Andrew Charlesworth is director of the information law and technology unit, University of Hull Law School.

(a.j.charlesworth@law.hull.ac.uk)

Who is likely to be caught by libel law?

While definitive answers await the attention of the courts, the answer to this appears to be:

*The poster of the defamatory statement to a bulletin board and potentially the system operator of the bulletin board;

*The poster of the defamatory statement to a moderated mailing list and the moderator, where the moderator reads all the messages for content;

*The creator of a defamatory Web page and potentially the owner of the Web server on which it is based;

*The creator of a link to a defamatory Web page where that link will have the effect of spreading the defamation, and where the creator of the link can be shown to have knowledge of the contents of the defamatory Web page.

Avoiding a libel action

Institutions should:

*Forbid the use of official logos or designations on personal Web pages.

*Disclaim any editorial responsibility for personal Web pages.

*Have a clear procedure in place for the removal of any material brought to its attention as potentially libellous.

*React swiftly and effectively when notified that potentially libellous material is being published via its machines.

This would usually entail removing the material, disciplining the creator of the Web page or email message, and offering a full apology to the individual concerned, in the media of their choice.