Washington Secretary of State Steve Hobbs is accused of using an AI program to carry out mass surveillance of citizens, i.e. voters in general, conservative activists, and also journalists – a way, critics say, to “weaponize” technology during the campaign leading up to US presidential elections.
The program is made possible by the UK company Logically, which says it’s in the business of producing ways to deploy “advanced AI to fight misinformation at scale” and is using fact-checkers, OSINT (so-called “open source intelligence) investigators, and “data science” to achieve this.
Logically’s previous “claim to fame” (and no doubt, recommendation as far as some politicians and political options in the US are concerned) is that the UK government used it to hunt down social media users there during the pandemic, if they were seen as “Covid misinformation spreaders.”
Controversial in and of itself, and already the subject of an ethics complaint, Washington’s decision to put such a program in action just got even more negative publicity after several journalists (reports mention Post Millenial’s Ari Hoffman and Katie Daviscourt) and leader of Republicans in the state Jim Walsh got caught in this surveillance net.
Given that their posts about the voting of non-citizens, criticism of automatic registration of voters, mail-in ballots, etc., were all factual, yet flagged by the program, the sense is that it is information and “narratives” the authorities such as Hobbs, a Democrat, don’t like – rather than “disinformation” – what Logically is “fighting at scale” here.
The explanation given for flagging the stories referring to the said issues’ in effect confirms it – they have the potential to “degrade confidence” in the state’s electoral process. And whether or not that confidence warrants “degrading” – due to irregularities – doesn’t seem to be a topic of consideration.
The big concern opponents have is how this stifling of speech “at scale” will affect the elections coming up this fall. Some of the services Logically provides to Washington State are monthly reports alerting about “threats” and (unwanted) “narratives” the surveillance system detects, and if an account is considered as threatening election officials, then Hobbs’ office gets informed about those and the people behind them.
Hobbs’ office, meanwhile, denies that information received from Logically gets used to “seek censorship, categorize specific social media users, or request action against any social media user.”
