Randomly generated tweet by bot prompts investigation by Dutch police
|
Version 0 of 1. When Twitter user @jeffrybooks tweeted saying “I seriously want to kill people” at a fashion and cosmetics convention happening at Amsterdam, the Dutch police took the threat seriously. Jeffry van der Goot, a 28-year-old web developer who created the account, received a visit from the police in short order. The only problem was that Van der Goot – who is non-binary and goes by “they” – hadn’t written the tweet. Instead, Van der Goot had handed the Twitter account over to “jeff_ebooks”, a bot they created which could automatically send tweets and hold conversations as an eerie simulacrum of themselves. It was jeff_ebooks who had sent the threat, and it was its creator who was being held liable. Van der Goot told the Guardian that it makes no sense that they would be hauled in for questioning over a fairly mild claim made by an automatic Twitter account. “I told [the police] that I can technically see that it would be my responsibility since I started the bot and it is basically tweeting under my ‘name’,” they said. “However, it is a random generator, so yes it is possible that something bad can come out of it, but to treat it as if I made that threat does not make sense to me. I feel very conflicted about it, I can see their point but it does not feel right to me that the random output of a program can be considered something I said.” The bot is what’s known as a markov chain generator, which uses a simple algorithm to create vaguely coherent sentences from a corpus of text. In this case, it uses a collection Van der Goot’s tweets as the basis for its own. “What it does is it takes fragments from the entire archive and creates new coherent sentences from the formerly unrelated sentence fragments,” they explain. To make matters even more confusing, the bot was engaged in a conversation with another bot, each having been programmed to reply to messages from strangers. “So they were just spamming random stuff at each other. I tried to clarify, but people seem to be under the misconception that the death threat was aimed AT the other bot, which is not the case.” At the police’s request, jeff_ebooks has been lain to rest, leaving Jeffry shaken by the whole thing. “Being interviewed by detectives is incredibly stressful, terrifying and intimidating,” they said. Wxcafe, the creator of the framework jeff_ebooks was based on, also expressed alarm, tweeting that they were “very sorry this happened”. Obviously they're automated and what they say is just based on the algorithm and the corpus, but like. Police involvment is... scary Of course since I don't have any legal knowledge I don't know who is/should be held responsible (if anyone) but like. kinda scared right now So yeah, hope it never happens again. But one positive highlighted by Van der Goot’s experience is that, in the Netherlands, police do at least take threats made on social media seriously – even when not made by real people. |