“Free software” activist wins libel claim over alleged responsibility for harassment campaign

Background

The claimant, Dr Matthew Garrett, is a software engineer and self-described “free software” activist, with a doctorate from Cambridge University. He is currently based in “Silicon Valley” in California, where he holds a senior position at NVIDIA and teaches part-time at the University of California, Berkeley.

The defendants, Dr Roy Schestowitz and Rianne Schestowitz, are a married couple based in the UK in Manchester. They run the informational websites Techrights and Tuxmachines respectively.  

In April 2024, Dr Garrett brought a libel claim in relation to 24 articles that appeared on the defendants’ websites between August and September 2023, as well as a data-protection claim. The defendants counterclaimed for harassment under section 1 of the Protection from Harassment Act 1997.

Before issuing the claim, Dr Garrett asked the defendants to remove the publications from the websites, but they refused to do so. They argued that none of the publications was likely to cause serious harm to the claimant’s reputation and asserted defences of truth, honest opinion and publication on a matter of public interest under sections 2 to 4 of the Defamation Act 2013. They sought to rely on the journalistic-purposes exemption for the data-protection claim.

The defendants represented themselves at trial. They did not serve any witness statement or witness summary for use at trial, maintaining that it was a deliberate decision for financial reasons. Yet it meant that they could not call any witnesses to give oral evidence at trial.

Preliminary issues

Meaning

Unusually for a libel trial, there had been no prior determination of meaning as a preliminary issue, as the parties’ pleadings had not been far apart in that respect.

So, Collins Rice J considered whether the natural and ordinary meaning of the words was defamatory, applying the established test, i.e. whether it has a substantially adverse effect on the attitude of other people towards a claimant, or has a tendency to do so. She also noted the relatively recent authority of Millett v Corbyn, ^[2] which frames the test in terms of identifying whether the claimant has breached the common, shared values of society.

When applying the law to the statements, the judge noted that her preliminary views on meaning had to be moderated in light of the meanings put forward by the claimant, including withdrawing a meaning where it went beyond what was argued by the claimant.  

Collins Rice J identified common themes repeated across the publications, which alleged that Dr Garrett was conducting a criminal, illegal or offensive online campaign. The allegations of criminality included cybercrime, hate crime, blackmail, threats and terrorism-adjacent matters, among others, such as abuse and harassment.

The judge found that the defendants had also alleged that Dr Garrett had conducted an online campaign relating to offensive matters, including homophobic, racist, antisemitic and misogynistic material. The campaign was said to have been waged by him on an internet relay chat through “sock-puppet” accounts, which operate under pseudonyms and distribute material anonymously. They repeatedly alleged that he was a habitual user of illegal drugs.

On the allegations of criminal or illegal acts, the judge agreed with Dr Garrett that they were of a Chase level 1, and that an ordinary, reasonable reader would have “no difficulty in understanding” that he was said to have committed the activity alleged. The judge determined the meaning of the allegations relating to offensive attitudinal matters to be closer to that asserted by the defendants, namely that they were expressions of opinion. The judge concluded that there was no doubt that each of the publications was of grossly defamatory tendency at common law, alleging “multiple and serious transgressions of society’s norms, up to and including the criminal law”.

Serious harm

Dr Garrett put forward a wholly inferential case on serious harm. The judge considered the legal framework and acknowledged, in light of the Court of Appeal decision in Blake v Fox, ^[3] that an inference of serious harm may be “irresistible” where claimants can establish that there has been wide publication, the words are gravely defamatory, and they have experienced significant reputational exposure.

The defamatory imputation of the words was determined by the judge at trial, and so she found no difficulty finding that the allegations were sufficiently serious for the purposes of serious harm.

Dr Garrett was born in Ireland and raised in the UK. Yet, as he emigrated to California in 2009, he had not lived in the UK for 16 years. So, one hurdle for him in proving serious harm was establishing that he had a suffered serious reputational harm in the jurisdiction of England and Wales. He argued that he had an established business and personal reputation in the UK that was “immensely important” to him, pointing to his work for the University of Cambridge and other work done in Cambridge, as well as his work for NVIDIA, which does business in the UK, and the fact that many of his friends and colleagues are in the UK. The judge accepted his evidence on this.

The main issue in dispute on serious harm was the extent of publication within the jurisdiction. No analytical data were available to the court for the publications, although it was accepted that the websites attracted around one million hits per day.

Counsel for the claimant argued that the websites had a readership “in the tens of thousands” for each article. By contrast, the defendants argued that around half of the million visits could be discounted as machine-generated and estimated the UK audience percentage as being in the low single figures, contending that the UK readership was not likely to be “above tens” on any given day. In particular, they pointed out that they did not publish on or use mass social media platforms, which necessarily meant that their viewing figures were smaller.

Dr Garrett, however, argued that: (a) the articles had been online for over two years; (b) some of the articles were published across both websites and hyperlinked to each other; (c) recent searches for his name on Google were returning links to the articles on the first page of results (indicating the continued accessibility and historic popularity of the websites); and (d) the websites were well established, and the defendants enjoyed a trusted reputation in the English-speaking free-software community.

He also sought to rely on Armory v Delamirie ^[4] as authority for the legal principle that, where there are gaps in trial evidence, the party responsible cannot benefit from them. Applied to this case, the defendants should not be able to rely on their failure to retain the analytical data or to provide any evidence of readership of the articles.

The judge found that the publications were accessible to, and accessed by, a section of the public in the UK looking for information about Dr Garrett, and that the websites were “popular and authoritative” among a free-software community in the UK of tens or hundreds of thousands of people. On that basis, the judge determined that an audience in the thousands was a fair estimate, and that the claimant could rely on the Armory principle on that point.

The judge held not only that a readership in the thousands could amount to a mass publication, but that the publication was made to precisely those individuals – people interested in the free-software community who were aware of or may be interested in Dr Garrett – with whom his reputation was most closely engaged. She noted that it was “hard to imagine a more serious assault on the professional and personal reputation of a prominent free software activist” than the allegations of criminality and imputations about Dr Garrett’s private life.

Accordingly, the judge accepted that the publications had caused (and were likely to cause) serious reputational harm to the claimant.

Defences

Truth

The defendants sought to rely on an inferential case, including a small number of messages that Dr Garrett accepted posting and his asserted authorship of a campaign of threats, abuse and harassment using the sock-puppet accounts. He denied authorship or control of those accounts, and the defendants advanced no evidence in support of their defence. In any event, the judge drew a distinction between the incident admitted by Dr Garrett and the “covert use” of sock-puppet accounts to publish illegal or offensive material.

The judge found that the defendants’ suspicions that the claimant had been behind a campaign of sock-puppet trolling experienced by Mrs Schestowitz were explained, but were “simply not capable of sustaining the weight of a truth defence”. Dr Garrett was described as a “straightforward witness” who provided a clear account. So, the truth defence failed in its entirety.

Honest opinion

For a defence of honest opinion to succeed, it must satisfy certain conditions, including that an honest person could have held the opinion on the basis of facts that existed at the time when the statement was made (section 3(4)(a) of the 2013 Act). The defence was pleaded for the allegation that Dr Garrett was responsible for harassing the defendants. Yet the defence failed, as the judge could not uphold the truth of the allegation on which the defence was based.

Public interest

The judge considered the three-stage test in Turley: ^[5]

1. Was the statement (or did it form part of) a statement on a matter of public interest?
2. Did the defendant believe that it was in the public interest to publish the statement?
3. Was that belief reasonable?

The judge accepted that the defendants’ observations about trolling obstructing promotion of high standards of online behaviour could fall within the broad ambit of a matter of public interest, and so could have met the first stage of the test. Yet they did not choose to write in general terms about their concerns along these lines, but instead focused “squarely on attacking” the claimant.

As the defendants advanced no evidence as to whether they believed the statement was in the public interest, there could be no assessment of the second or third stages of the test, i.e. as to whether they believed publication to be in the public interest, and whether that subjective belief was objectively reasonable.

The judge noted that, even if the evidence had been provided, the defendants would have faced difficulties in demonstrating that they took appropriate steps prior to publication, as there was no evidence that they had conducted any due diligence that might support an objective reasonable belief. In particular, the defendants did not approach Dr Garrett to give his side of the story.

In light of those issues, the public-interest defence also failed.

Conclusions

The judge acknowledged the traumatising effect on the defendants of the campaign of abuse and noted that their websites were well regarded in the free-software community. Yet they had publicly and repeatedly blamed Dr Garrett in a way that was “bitter, highly personal and thoroughly derogatory”.

By not providing any evidence to substantiate the defences they advanced, the judge was obliged to find for the claimant. The defendants’ counterclaim for harassment failed on the basis that they had not established the claimant’s responsibility for the abuse that they received.

In assessing Dr Garrett’s claim for damages, the judge noted that the gravity of the allegations, published on reputable websites to a substantial audience in which the claimant’s reputation was most deeply invested, profoundly impugned his personal, social and professional standing.

The claimant was awarded damages of £70,000 in his libel claim, and the court ordered the defendants to publish a summary of the judgment on their websites under section 12(1) of the 2013 Act.

Comment

The judge observed that, rather than seeking to raise concerns in an objective manner over the highly distressing abuse and harassment they received, the defendants alighted on Dr Garrett as being responsible for that online campaign, and they used their websites to make serious accusations that caused grave damage to his reputation. They chose to defend the claim without having adequate evidence that he was responsible for the harassment that they had suffered. Further, even if the truth of the allegations could not be shown, they had no other defence available to them where they had failed to take the proper steps required of publishers to satisfy themselves that publication was in the public interest.

The judgment underlines the risks of publishing unsubstantiated allegations online, particularly by individuals without journalism training and so apparently lacking a proper understanding of the legal risks and pitfalls. The websites had niche but engaged audiences that were relevant to the claimant’s professional engagement with the free-software movement. So, the potential for reputational harm to him was significant, given the nature of the community served by the publication in question. That was borne out in the size of damages awarded.

‍

Article written for Entertainment Law Review.