Researchers identify over a thousand fake LinkedIn accounts with computer-generated profile pictures being used for lead generation and product promotion
Loading... At first glance, Renée DiResta thought the LinkedIn message seemed normal enough.
NPRShannon Bond
Context & Ripple Effects
Renée DiResta's discovery of over a thousand computer-generated faces running lead-generation pitches was an early field sighting of what has since become LinkedIn's defining integrity problem. Months later the same technique had graduated to fake executive profiles pairing AI photos with text lifted from real accounts, enough of a nuisance to hit HR departments and invite-only groups.
Anyone targeted by these outreach messages — sales prospects and recruiters above all — is being sold to by personas that do not exist, and the product being promoted gets distribution LinkedIn's ad system never approved.
Second-order effects
LinkedIn absorbs a ballooning enforcement cost, visible in its own transparency numbers, while legitimate users and HR teams pay a screening tax on every new connection request and invitation from an unfamiliar name.
Third-order effects
If synthetic personas keep scaling, professional networking drifts toward an authentication arms race — verified identity becoming the scarce asset, with fake-account operations eventually treated as a legal problem too, as LinkedIn's later suit against ProAPIs over alleged fake-account-driven data scraping suggests.
The trend: Professional networks are entering a synthetic-supply era where AI-generated people, posts, and profiles force platforms to shift spending from content moderation toward identity verification.
Meet Keenan Ramsey. Her LinkedIn profile says she sells software for RingCentral & has a business degree from NYU. She likes CNN, Amazon, & Melinda French Gates. Her pitches come punctuated with emojis. https://www.npr.org/... https://twitter.com/...
Welcome to the latest LinkedIn marketing tactic: using AI-generated faces to drum up sales. A tool used to spread disinformation has come to the corporate world. Stanford's @noUpside & @JoshAGoldstein found >1,000 profiles with seemingly fake faces: https://www.npr.org/... https:…
Really excellent piece by NPR investigating some of the behind-the-scenes dynamic of a network of 1000+ fake accounts on LNKD that @JoshAGoldstein & I found a month or so ago. The find started by accident - this random account messaged me... https://twitter.com/...
And people on social media use faces from LinkedIn to appear like users with specific demographic characteristics while spreading disinfo... https://twitter.com/...
Wow .. but when #yourslipishowing asked these people to cover this it was nothing Folks are more enamored of AI fake people than protecting actual Black people https://twitter.com/...
Well well well ... the facts finally don't make LI look any good or “professional” a word they crow about hugely. This is not merely disgusting but treacherous as well. https://twitter.com/...
For every adversarial defense there seems to be an attack that breaks it. In this case, NPR advises looking for asymmetry and centered eyes as telltale signs of fake social media photos, but a small modification the the GAN loss or generator would make this “defense” useless. htt…
Researchers at Stanford uncovered more than 1,000 LinkedIn profiles using what appear to be faces created by artificial intelligence. A technology that has been used to promote misinformation and harassment online has now entered the corporate world. https://www.npr.org/... https…
“We call it ‘narrative collapse’. When a planet is very connected, a time inevitably arrives when it becomes difficult to work out what is actually going on. Video and audio can be faked. Testimony isn't reliable. All truths fall into a relative flatness.” ~"The Fifth Science" ht…
My story just scratches the surface of a new evolution in fake accounts. If you work at a company that's used AI-gen faces for sales (or anything else!), or a company that sells LinkedIn marketing services, I'd love to hear from you. DMs are open. https://www.npr.org/...
I had such a profile showing up as an employee of @TantraAnalyst on @LinkedIn . It took >2 weeks of back n forth with #linkedin to remove the association. They declined to remove the profile though. #cybersecurity https://twitter.com/...
Given that, it's funny this whole story came about bcs that Keenan Ramsey profile sent a LinkedIn message to @noUpside — one of the few people who CAN spot the telltale signs of an AI-generated image. “The face jumped out at me as being fake,” she says. https://www.npr.org/... ht…
Unreal. Identifying the #AI-generated faces used to create fake identities and spread misinformation on social media. (FYI @ProfXtian) https://twitter.com/...
But...RingCentral doesn't have any record of an employee named Keenan Ramsey. NYU says no one named Keenan Ramsey has received any undergraduate degree. And the biggest red flag? Her face appears to have been created by artificial intelligence. https://www.npr.org/... https://twi…