Researcher Rick Claypool couldn’t help but draw on the Pinkerton analogy when he discovered that Feedly the news aggregator apparently launched artificial intelligence models that let users “keep up with protests, riots, and strikes” that are “posing a risk to your company’s assets.”
The models, called Protests and Violent Protests use AI to crawl “millions of sources” on the web for trends that suggest emerging protests based on location, events, companies, and specific people. It’s a paid service that the researcher pointed out entails “surveilling workers and activists on behalf [of] corporations.”
As Claypool notes, the models were launched earlier in March via Twitter, people just didn’t notice.
When they did, Feedly took back its comms, deleted the posts, and blamed it all on an unclear messaging on the blog post, telling Claypool that the models are “used to notify employees of a potential security risk if they travel or work at locations where violence emerges. The content was misleading, so deleted.”
The now-deleted blog post that he shared was published by Feedly AI on March 29. That, and a screenshot of the now-deleted tweets from March 15 clearly say that the tracking models were designed for companies and to help them protect their assets.
Feedly was initially released in 2008 and was, for a long time, the aggregator that people used to track their favorite publications and blogs. Feedly AI offers a collection of paid AI models that track threat intelligence, market intelligence, risk intelligence, and biopharma research.
Information for this briefing was found via Rick Claypool on Twitter, and the sources and companies mentioned. The author has no securities or affiliations related to the organizations discussed. Not a recommendation to buy or sell. Always do additional research and consult a professional before purchasing a security. The author holds no licenses.
