Kellogg Insight: Is There a Bot Behind That Tweet?. “When we see messages that contradict our political ideology, we are more inclined to attribute them to bots. It’s making society even more polarized.”
The Guardian: Elections in UK and US at risk from AI-driven disinformation, say experts. “Next year’s elections in Britain and the US could be marked by a wave of AI-powered disinformation, experts have warned, as generated images, text and deepfake videos go viral at the behest of swarms of AI-powered propaganda bots.”
Rolling Stone: Twitter Bots Are Promising Cheap Guns to Anyone Using the N-Word. “When Elon Musk acquired Twitter, he made it clear that eradicating spam bots was a top priority. But his strategies on that front have produced, at best, mixed results. Moreover, in recent months, some users have been plagued by a type of spam that directly violates Twitter policy: automated replies from accounts offering guns for sale.”
Spotted on Mastodon: a Google Sheet showing Twitter bots moving to Mastodon. Since Twitter is making its API a paid service, a lot of these bots will stop working on Twitter soon. There are over 110 listings at this writing.
Curbed NY: I’ll Miss You, Design Bots. “Twitter bots like @slam_decorative crawl public records and public collections, tweeting historic, anarchic, and beautiful images and text into your everyday feed. They provide an escape from the influencers, an escape from the algorithm, an escape from 2023 — or they will until February 9.”
BuzzFeed News: Thousands Of Fake Twitter Accounts Were Made To Support Andrew Tate. “Escaping content featuring misogynist influencer Andrew Tate can be hard on social media — and according to new research, that’s because thousands of accounts posting favorable content about him are actually bots. The Center for Countering Digital Hate announced on Wednesday that it had found 4,621 fake Twitter accounts created to promote Tate. The bot accounts tweeted 15,202 times in support of the former kickboxer, who is in pretrial detention in Romania.”
Bloomberg: Musk’s Frequent Twitter Polls Are at Risk of Bot Manipulation. “New Twitter Inc. owner Elon Musk has outsourced several controversial decisions — like whether to reinstate former US President Donald Trump’s account, and if he should leave the Twitter CEO job — to public polling on the network, saying he intends to follow the will of the people. But the results of such surveys can be easily gamed by bots, according to new research.”
Atlanta Business Journal: Atlanta homebuilding giant PulteGroup fires incoming COO for alleged Twitter bots. “The founding family of the Atlanta-based homebuilder (NYSE:PHM) filed a civil complaint against Brandon Jones for ‘interfering, stalking, harassing, and defaming the company’s founder, William J. Pulte, his grandson Bill Pulte, and his family,’ according to a news release Friday. The complaint alleges Jones, who was set to succeed executive vice president and chief operating officer John Chadwick on Jan. 1, creating fraudulent Twitter profiles.”
KQED: This Bot Posts Rejected License Plates and It’s Amazing. “Each… post includes the applicant’s stated explanation for their license plate, the DMV’s concerns, the decision to accept or deny, and a rendering of the plate itself. If this sounds like boring government bureaucracy, may I direct you to drivers’ valiant, creative or altogether stupid attempts at deceiving the almighty DMV censors.”
Ars Technica: Twitter pranksters derail GPT-3 bot with newly discovered ‘prompt injection’ hack. “On Thursday, a few Twitter users discovered how to hijack an automated tweet bot, dedicated to remote jobs, running on the GPT-3 language model by OpenAI. Using a newly discovered technique called a ‘prompt injection attack,’ they redirected the bot to repeat embarrassing and ridiculous phrases.”
Pittsburgh Jewish Chronicle: A Twitter ‘Repentance Bot’ teaches people how to apologize for real. “‘Repentance Bot’ allows users to tag the account when they see an apology that they believe falls short. The bot then replies to the apology with encouragement to do better and a comic strip laying out five steps to take to do so.”
LA Taco: A New Twitter Bot Will Tell You An Officer’s Name, Rank, And Race By Putting In Their Serial Number. “WhosThatCop is a Twitter bot created by user @NN35007 that allows users to enter the serial number of an officer they’ve encountered, which then reveals the name of the officer, their job, gender, and race.”
The Print: 60-80% of Twitter accounts posting on Russia-Ukraine war bots, 90% ‘pro Ukraine’, finds new study . “Between 60 and 80 per cent of Twitter handles posting on the Russia-Ukraine war may be bot accounts, a research by scholars from the University of Adelaide, Australia has found. Among other influences, these bot accounts may have been pushing people to flee their homes during the conflict between these two countries, the researchers added.”
New-to-me, from Poynter: Repustar debuts a tip line tool for US fact-checking organizations. “Repustar, a fact-checking organization that crowdsources fact checks, is partnering with journalism and fact-checking outlets in the US to provide both fact checks to the public and verifiable claims to fact-checkers. It aims to rollout the service globally in 2023. The remotely operated startup is also responsible for FactSparrow, a Twitter bot that users can tag under tweets with claims they would like to be checked, and The Gigafact Project, a fact-checking editorial platform.”
Yahoo Finance: Elon Musk demands Jack Dorsey turn over Twitter info on fake accounts. “On Monday, attorneys for Musk filed court documents showing they had issued a subpoena requiring Dorsey to share documents on Twitter’s fake accounts, an issue at the heart of Musk’s rationale for backing out of the $44 billion agreement to acquire the company.” Jack Dorsey stepped down as Twitter CEO last November and completely left the company in May, so I’m not sure what this is trying to accomplish.
