ParityBOT_NZ Processed Nearly 200,000 Tweets To Find And Fight Online Abuse Over The NZ Election
Auckland, New Zealand: Of the almost 200,000 tweets directed at women candidates that Areto Labs’ proprietary election Twitter bot, ParityBOT_NZ, processed over the New Zealand election period, nearly 4,000 were classified as toxic and triggered the bot to send out its own “positivitweet”.
“Women in politics experience disproportionately more, and worse, abuse online. Fighting this is key to achieving gender parity. Deploying ParityBOT_NZ for the general election was a success: we were able to systematically tackle the worst tweets with positivity, and it made a difference,” said Lana Cuthbertson, CEO of Areto Labs.
Over the 10 weeks that ParityBOT_NZ ran up until election day, the highest number of toxic tweets in one day was 95, on one of the debate nights. Then election night blew that number out of the water: the increase in ParityBOT_NZ activity was ten-fold, at 1032 toxic tweets.
“We were really surprised with the exponential increase in toxicity on election night,” continues Areto Labs’ creative technologist and Aucklander Jacqueline Comer. “I knew something was up when I was trying to focus on the results coming in but my phone was buzzing non-stop with Twitter notifications.” While Kiwis were responsible for some of those bot-triggering hate tweets, many of them came from outside of New Zealand, and were in hateful response to world leaders like Justin Trudeau, Boris Johnson and Narendra Modi congratulating Jacinda Ardern on Twitter.
Comer: “It was really a mixed bag of toxicity, and basically the worst aspects of the platform culminating into one night of overactivity. And like moths to a flame, it seems that those peddling online hate are also attracted to any spotlight.”
In fact, throughout the extended election campaign, the bot’s nearly 4,000 tweets were mostly directed at high-profile candidates and party leaders, including Jacinda Ardern (1), Judith Collins (2) and Chloe Swarbrick (3). Swarbrick was also the receiver of a potential coordinated swarm of similarly-worded toxic tweets on election night.
“Research has shown that high profile women tend to attract more hate online--we saw this last year in Canada as well, and we’re seeing it in the U.S. right now. And while abuse during elections is high, it just gets worse once women are elected and take office,” said Kasey Machin, COO of Areto Labs.
Areto Labs is keen to keep ParityBOT_NZ running, and is looking for one or a few partner organizations or individual champions in New Zealand to contribute to the effort.
“Now that we have the bot built and running, we just need a minimum investment to maintain it. Anyone who wants to see this effort continue can contribute even $5 a month on our Patreon page,” Machin said.
Designed and launched by Canadian and NZ women-owned tech company, Areto Labs, ParityBOT_NZ uses artificial intelligence to classify hateful tweets sent to women candidates and posts positive tweets in response, to raise awareness about the issues faced by women campaigning online, and build support for candidates when they’re faced with online abuse. Areto Labs launched ParityBOT_US in September 2020 and has plans to launch other equity bots around the world in the near future.