Twitter CEO Costolo "Ashamed" of Abuse

Dick Costolo, CEO Twitter.

In a leaked memo, revealed in The Verge, Dick Costolo, CEO of Twitter (@dickc) has admitted he is ashamed of the way Twitter deals with abusive trolls and fails to protect it customers.

But he needs to do more than just control the trolls if he wishes Twitter to remain a platform for free speech and for exchanging ideas, no matter how they might upset vested interests, people who simply don't like people disagreeing with them and think it shouldn't be allowed, or people exposed for lies, deceptions and scamming gullible and vulnerable people who currently infest the Internet.

I'm frankly ashamed of how poorly we've dealt with this issue during my tenure as CEO. It's absurd. There's no excuse for it. I take full responsibility for not being more aggressive on this front. It's nobody else's fault but mine, and it's embarrassing...

It’s no secret and the rest of the world talks about it every day. We lose core user after core user by not addressing simple trolling issues that they face every day.

Dick Costolo

This frank and welcome confession was made on a Twitter internal forum in response to an employee who had drawn attention to a Guardian article by comedian and feminist, Lindy West in which she talked of the deluge of abuse and obscenities, including setting up fake accounts to do it through (sound familiar?) to which she was constantly subjected - not just on Twitter but on other social media and in emails too.

There's a term for this brand of gratuitous online cruelty: we call it internet trolling. Trolling is recreational abuse – usually anonymous – intended to waste the subject's time or get a rise out of them or frustrate or frighten them into silence. Sometimes it's relatively innocuous (like asking contrarian questions just to start an argument) or juvenile (like making fun of my weight or my intelligence), but – particularly when the subject is a young woman – it frequently crosses the line into bona fide, dangerous stalking and harassment...

Social media companies say, "Just report any abuse and move on. We’re handling it." So I do that. But reporting abuse is a tedious, labour-intensive process that can eat up half my working day. In any case, most of my reports are rejected. And once any troll is blocked (or even if they’re suspended), they can just make a new account and start all over again. [My emphasis]

Lindy West

I can appreciate that 'most of my reports are rejected' comment, having variously reported explicit death threats (DEATH THREATS!) and threats to harm my family from Christian fundamentalists, only to be told by Twitter Support that they've asked the offender not to do it again but won't be taking any action; repetitive defamatory accusations of fraud and paedophilia, to be told the account isn't currently breaching 'Twitter Rules'; claims to be revealing my private name and address (falsely, but Twitter werent to know that) to be told again that this doesn't breach Twitter Rules (which explicitly forbid revealing private information without consent); examples of ongoing abuse and harassment to be told the account has been deleted when it is still active and tweeting abuse.

As I've commented before in this blog, although I acknowledge that there seems to have been a recent attempt to improve things, Twitter Support often appear to select a random response to a complaint, including 'the account has been suspended' when it hasn't been. As I've shown, reporting the same tweet again can produce a different response, from, 'have found that it's currently not violating the Twitter Rules', 'We've contacted this user and asked them to discontinue this behavior' and 'we have suspended the account'. Clearly the very same Tweet can't be grounds for suspension for abusive behaviour, grounds for asking the account holder to discontinue his behaviour, and not be in breach of Twitter Rules, all at the same time.

But this is not the only area in which Twitter need to improve their service.

Twitter is notorious for suddenly suspending perfectly innocuous, if controversial, accounts with no history of abuse or personal attacks, as it did recently with @GSpellchecker. This seems to be following a concerted campaign of false 'spam blocking', where an account is blocked and reported for spam, which has triggered an automated 'algorithm' suspension. After a couple of these, the account is normally suspended permanently. The only reason you will be given is a list of possible reasons, but no examples of offending tweets will ever be provided (on the curious grounds that this would be a breach of confidence - the reason my account was suspended is confidential?!).

Even when presented with unambiguous evidence that this is going on in the form of tweets explicitly appealing for it to be done and linking to instructions for how to do it, Twitter Support currently take no action because this doesn't currently breach their Terms of Service. I have even known this process to be automated so Twitter is bombarded with false spam blocks, and claims that you can get an account suspended within minutes if you use this 'service'.

With all that in mind, I posted the following as a comment to the Guardian article about this leaked memo:

At last they're acknowledging that they handle things badly. What they also need to address is the form of abuse which allows effective censorship by a concerted campaign of false reporting for 'spam', which appears to be fully automated so your account can be suspended because so many people disagree with your views and want you shut up.

This happens frequently, particularly to those expressing Atheist or Humanist views and views opposing Creationism and Bible literalism, and maybe other controversial but legitimate views too, so a perfectly polite, thoughtful or witty account can suddenly be suspended for 'breaching the Twitter TOS', although no offending tweet can ever be be produced.

It happened recently and most famously to @GSpellchecker but in this case, following a media campaign supported by a few prominent individuals, the account was 'reviewed' and reinstated. No examples of offending tweets were produced and Twitter hid behind 'breach of privacy'.

No such courtesy is extended to others who don't manage to get media publicity embarrassing to Twitter and there appears to be no formal appeals mechanism. One is merely confronted by an impenetrable wall of anonymous Support (sic) with it's stock replies and refusal to reconsider, as I've found to my costs.

In effect, Twitter have abdicated their responsibility in this area and handed it over to the abusers who wish to suppress free speech and shut down dissent. I recognise that they need to control advertising spam and that reporting it where appropriate is legitimate, but Twitter should also apply sanction to those who abuse this facility by suspending accounts making false or spurious reports, discontinue their policy of automated 'algorithm' suspensions, provide evidence of any breaches of TOS where requested and instate a formal appeal procedure.

As things stand the abusers will continue to have the upper hand and Twitter will continue to dance to their tune.

Please feel free to add your comments to mine.

Obviously it immediately attracted the attention of an idiot who seemed to take exception to the fact that I had mentioned that 'polite' accounts can find themselves suspended and who then went on to complain about 'polite' tweets criticising the Egyptian government as examples of abusive trolling. I guess the Internet attracts some strange people.

I wonder if, as part of his taking personal responsibility for the poor way Twitter deals with abuse, Dick Costolo might also look at the way accounts are suspended without the right to an appeal and without providing examples of offending tweets, and how Twitter can recapture control of this from the abusers to whom they have effectively handed control and abdicated their responsibilities. Censoring accounts because you disagree with them, and lying to get them suspended, is also a form of abuse with which, at the moment, Twitter seems to be cooperating and even facilitating because it's the cheap and easy option.

At the very least, Twitter needs to be suspending accounts registering false spam reports and unfounded complaints rather than suspending the victims of this abuse.

---