Twitter Abuse: Law and Myth
The first words of this piece were written while listening to Caroline Criado-Perez and Steve Kuncewicz on Radio 5 live on Sunday 28 July, discussing the multiple threats of rape and violence directed at Criado-Perez after her successful campaign to put a woman on English and Welsh banknotes1.
Criado-Perez is one in a long line of female campaigners who have suffered a violent tirade of on-line abuse after speaking out on sexism. Depressingly, the response to Criado-Perez, as to others such as Anita Sarkessian and Rebecca Watson, has generated more heat than light.
In part, this may be because of prevalent myths about the legal status of Twitter (and other online platforms). These myths include:
- Twitter already has effective systems in place to deter abuse
- Twitter acts only as a conduit and is not liable for its users' acts
- What is being proposed creates a legal right which did not previously exist
- What is being proposed is a one-stop, automatic ban function
- Abuse is a basic reality of life on-line; anyone bothered by it should leave
- Proposals to make abuse reporting easier attack freedom of speech.
A recent article in the Telegraph on-line by the newspaper's tech blogger Mic Wright2 about reaction to the attacks on Criado-Perez enshrines a fair number of the above.
First, Wright's faith in Twitter's internal procedures to deal with abuse seems misplaced to the point of naiveté:
Twitter has a policy on abuse. Its UK general manager, Tony Wang, said today: “We don’t comment on individual accounts but we have rules which people agree to abide by when they sign up to Twitter. We take online abuse seriously and provide advice and guidance to our users … also, we're testing ways to simplify reporting, e.g. within a Tweet by using the "Report Tweet" button in our iPhone app and on mobile web. We will suspend accounts that, once reported to us, are found to be in breach of our rules.”
Wang's promised changes to the technical process of reporting abuse cannot come soon enough. Finding the "Report Abuse" form on the Twitter site at present is reminiscent of searching an abandoned office building for the locked filing cabinet which may or may not be behind the door marked "Beware of the Leopard".
First, Wang's promise to suspend "accounts that, once reported to us, are found to be in breach of the rules" would be virtually useless unless the current rule permitting users to create new accounts when they have been suspended is also changed. The reason given for this restriction is difficulty of proof, but nothing is said about how to deal with the problem where it is not merely clear a suspended user is making use of a second account, they actually use it to boast of flouting the rules4.
A further major difficulty with relying on Twitter's own procedures comes with the section headed "Who can report abusive behaviour on Twitter?"
In order to investigate reports of abusive behaviours, violent threats or a breach of privacy, we need to be in contact with the actual person affected or their authorised representative. We are unable to respond to requests from uninvolved parties regarding those issues to mitigate the likelihood of false or unauthorised reports. If you are not an authorised representative but you are in contact with the individual, encourage the individual to file a report through our forms.
Wright is concerned that any changes to this policy might mean that
A Twitter user with many thousands of users could quite easily encourage their fans to click it to silence a critic they disliked.
Those terms include the following:
Unlawful Use: You may not use our service for any unlawful purposes or in furtherance of illegal activities. International users agree to comply with all local laws regarding online conduct and acceptable content.
As it stands at present that principle sets out a rule without a remedy. There is currently no mechanism for reporting a breach of law otherwise than by reporting it as an "abuse". The requirement that "abuse" may only be reported personally by the victim or their representatives builds in a fatal level of delay in getting Twitter's attention to serious and unfolding legal problems.
For example, those using Twitter to name the teenager raped by footballer Ched Evans committed an offence under relevant sexual offences law, as well as contempt of court. However, anyone attempting to report them to Twitter was met with a form email refusing to entertain a complaint about behaviour which was clearly a breach of law and a TOU breach unless they could prove they were either the rape victim or her legal representative.
Twitter's timidity in failing to shut down the accounts of those responsible caused the behaviour to go on far longer than it need have done. Although a number of the perpetrators were eventually successfully prosecuted, this was in spite of Twitter, not because of it. A substantial number of perpetrators escaped scot-free and continue to tweet unmolested. By contrast, the young woman who was the victim of the crime has been forced to change her name and move from the vicinity.
When, as happened to Criado-Perez, people are tweeting rape threats accompanied by what they believe is the home address of the person threatened matters have gone beyond "abuse" into criminal incitement. Twitter must adopt some way of dealing with clearly criminal behaviour which is quick, effective, and does not require the target to respond personally. After all, they may have enough on their plate with changing the locks and speaking to the police than have time for burrowing through the arcane complaints procedure. They may not even have a Twitter account.
Wright also seems to be confused about how much responsibility Twitter has under current law:
To some extent, holding Twitter so directly responsible for abuse that flows through its network is like demanding that the Post Office inspect every letter.
Not to a very great extent, actually. Under a legal regime going back over 150 years, the Royal Mail has special status as a common carrier. It even enjoys a specific exemption under the "communications data" provisions of the Regulation of Investigatory Powers Act 2000 (RIPA). This sets it aside from organisations which operate as "electronic communications service networks" who are already obliged to give subscriber information or traffic date to public authorities requesting it under RIPA and who will be subject to even more extensive powers if the revived Communications Data Bill becomes law.
Under RIPA, almost six hundred thousand notices and authorisations to access communications data were, in fact, granted in 20125, a fact of which Wright — when referencing "countries where genuine free speech is curtailed" — seems to be unaware.
In any case, Twitter is not a "mere conduit" as Wright seems to assume. By the standard set out in s.32(7) of the Communications Act 2003 6 it is a "content service" in that it exercises a level of editorial control.
Undoubtedly, the issue of on-line freedom of speech, surveillance and on-line snooping is a genuine concern. However, it is something which needs to be looked at as something which affects all users of electronic communications networks, not something wheeled out to combat protests about abuse directed at anti-sexism campaigners, and ignored the rest of the time.
Ultimately Wright is correct when he states
But abusive words aren’t a technology problem: they are a societal issue.
Societal problems can, however, have technological solutions. Furthermore, Twitter cannot claim to be independent of the society whose crimes it currently facilitates. While "abuse" is a subjective term, a "crime" is something which has been defined as such by statue. Relevant statutes include the Protection from Harassment Act 1997, s. 127 of the Communications Act 2003 and the Sexual Offences Act 2003. Any criminal conduct report form should require the complainant to indicate the nature of the crime alleged, to set it apart from plain "abuse".
Much is often made of the presumption of innocence in these cases. However, no-one would ever criticise the manager of a supermarket who caught someone on CCTV apparently shoplifting for reporting them to police, notwithstanding that the presumption of innocence would apply once the case was brought to trial. The same approach is all that is being requested of Twitter.
Scare stories by freedom of speech campaigners suggest that what is being proposed is an automatic report-and-ban function which would be bound to be abused against minorities. These arguments both ignore the realities of what is being proposed and also overlook the chilling effects which already exist on the speech of women who are routinely driven off social media and deterred from campaigning by the unchecked abuse and threats which now inevitably follows. Furthermore, they carry the implication that where campaigners disapprove of relevant law, the answer is not a campaign to repeal the law, but selective non-enforcement if it.
It is certainly true that there has been ham-fisted enforcement of law relating to social media in the past; the ill-fated "Twitter joke trial" led to the formulation of new CPS guidelines on prosecution by the CPS7. However, anyone reading the tweets directed at Criado-Perez can have little doubt that a substantial number of them fall squarely within paragraph 12 of those guidelines. Terms like "abuse" and "trolling" as used by writers such as Wright are euphemisms for sustained, coordinated criminal conduct.
Accordingly, what Twitter is being invited to do is to consider more effective ways crime can be reported to it.
With Wright, I remain sceptical about the effectiveness of the proposed #trolliday from Twitter (August 4th) to mark opposition to threats of this sort. Instead, I propose an alternative. Women using Twitter on 4th August should add the hastag #internationalairport to remind Twitter that it historically has taken some criminal threats seriously and that the answer to fears of misplaced enforcement of criminal law is not no legal enforcement but better directed enforcement.
There is now a follow up post on Twitter abuse and contract law
This occurred a number of times during the furore about naming the Ched Evans rape victim as well as in the current Criado-Perez furore; in any event an argument based on practicality should not be confused with an argument in principle. Other service providers do indeed make it clear that deliberate attempts to circumvent a ban will lead to increased penalties. ↩
With respect to any prosecutions for offences disclosed, the CPS guidelines would still apply. CPS Guidelines on prosecution of offences involving social media ↩
The main image used for this article is: 'Jane Austen' and was used under the terms detailed at the above link on the date this article was first published.Tweet your Comments...