Maggie Harrison Dupré

I kept digging into that AI company behind Sports Illustrated's fake writers. Turns out it's created tons more fake writers across the web, sometimes for surprisingly huge media clients. And then things get even wilder... For @futurism // futurism.com/advon-ai-conte…

⚠️ "Many ChatGPT users are developing all-consuming obsessions with the chatbot, spiraling into severe mental health crises characterized by paranoia, delusions, and breaks with reality." - by @mags_h11 for @futurism 👉 futurism.com/commitment-jai… #ChatGPT #MentalHealth

NEW: We investigated several stories of people who spiraled into severe mental health crises after developing all-consuming obsessions with ChatGPT. Screenshots we reviewed show the AI directly engaging with and supporting user delusions/conspiracies. futurism.com/chatgpt-mental…

How "AI slop farmer" Jesse Cunningham targets older women on Pinterest and Facebook with AI-generated copies of content from real bloggers and creators (@mags_h11 / Futurism) futurism.com/slop-farmer-ai… techmeme.com/250508/p11#a25…

1/ 🧵 New preprint out now: "Delusions by Design? How Everyday AIs Might Be Fuelling Psychosis (and What Can Be Done About It)" We explore how LLMs may shape thinking in people vulnerable to psychosis — for better or worse. 📄 osf.io/preprints/psya…

Are Google AI Overviews leading you to the best results on the web? Or is this just a new prime spot for selling you products you will regret buying? Well, it turns out this rabbit hole is a deep one. housefresh.com/beware-of-the-…

NEW: People are being involuntarily committed to psychiatric care facilities — and even ending up arrested or jailed — after becoming fixated on ChatGPT and other chatbots and spiraling into psychosis. futurism.com/commitment-jai…

People Are Becoming Obsessed with ChatGPT and Spiraling Into Severe Delusions. “What these bots are saying is worsening delusions, and it's causing enormous harm." Great reporting from @mags_h11 futurism.com/chatgpt-mental…

Excellent reporting here by @mags_h11 abt how a sycophantic, enabling #ChatGPT is reinforcing unstable minds & sending people into deep spirals. I believe this phenomenon to be true & have seen evidence of it. futurism.com/chatgpt-mental…

NEW: SoundCloud has (again) updated its AI policies following backlash to a vague clause that allowed for the use of artists' music to train AI. Its CEO conceded that the language was "too broad." The updated policy requires opt-ins for genAI training. futurism.com/soundcloud-ai-…

Gannett is using AI to generate automated lottery result articles across local papers, often with vague bylines and affiliate links to lottery site Jackpocket (@mags_h11 / Futurism) futurism.com/gannett-ai-gam… mediagazer.com/250512/p16#a25…

Big, insane @mags_h11 story this morning unfolding in real time: SoundCloud updated its terms of service to allow AI to train on EVERYTHING ANYONE UPLOADS, and its users are just finding out now. futurism.com/soundcloud-ai-…

NEW: We dug into the tactics of an AI slop farmer, who uses AI tools and fake "bloggers" to flood social media with synthetic copies of competitors' content. His ideal audience, he's said, is older women, who may not know they're clicking on AI. futurism.com/slop-farmer-ai…

Wild story that shows how even student news outlets can be targeted by scammers who publish AI slop, thanks in this case to a lapsed domain name at Emerson's student-run radio station.

NEW: Since 2023, student journalists at Emerson College's student radio station, WECB, have been battling a digital imposter: an AI zombie site using a stolen WECB web domain -- and the station's likeness -- to churn out Google-ranking AI slop for profit. futurism.com/emerson-colleg…

Quick guardrail check on CharacterAI -- corroborating tests conducted in this risk assessment, we were easily able to get AI Wario, of Super Mario Bros. fame, to produce a recipe for napalm. *In Wario voice.*

NEW: In a new AI risk assessment, researchers warn that no one under 18 should be using *any* social AI companions — including CharacterAI. Potential risks to kids include bots encouraging destructive behaviors and exacerbating mental health conditions. futurism.com/stanford-no-ki…

NEW: For a decade, a man battled a severe, life-altering GI illness that doctors struggled to diagnose. Finally, he tried something new: he swallowed a PillCam — which wound up stuck in his intestines for months. You won't believe what it found. futurism.com/neoscope/filmm…

It's like throwing water at a lit fire when we need to prevent the fire from happening at all. If they can't take preventative actions, and obtain consent when needed, they need to rethink their business practices.