Weighing mainstream and alternative accounts…
Weighing mainstream and alternative accounts…
Rabbit hole
AI Fakes Are Slipping Through
A polished pitch lands in an inbox, a fake paper enters a journal, a harmful tool passes as an ordinary app, and a forged identity clears a financial check. The same weakness keeps showing up: systems judge what something looks like, while the machinery behind it stays out of sight. Elections and cyberattacks bring the danger home, because once imitation reaches a trusted channel, doubt spreads faster than any correction.
8 connections · November 5, 2026
Source mix
17 of 87 institutional
The ordinary storefront becomes the disguise
OnlyFake and the hidden deepfake apps turned routine-looking interfaces into cover for what happened after the gate opened. The trick was not to defeat every safeguard, but to present a harmless face at the moment of inspection.
Synthetic words are entering trusted channels
Automated pitches and fabricated scientific work exploit the same gap between fluent language and genuine authorship. Journalists, editors and detectors can notice clues, but a convincing surface still moves faster than a reliable answer about who made it.
Once the imitation passes, trust pays the price
AI-assisted attacks show how operators can use familiar tools to reach financial systems, while synthetic political material makes official messages compete with manufactured voices. The damage is not only stolen data or a false claim, but the growing suspicion that no trusted channel is what it seems.
AI-written press pitches are flooding newsrooms and often mask scams, a growing number of editors warn.
DocumentedEditors assigned and published work that later proved AI-generated or fraudulent, showing AI imitations bypassed newsroom checks and were only caught after publication or further scrutiny.
“We retracted a story in May.”How WIRED Got Rolled by an AI Freelancer | WIRED · www.wired.com →
“You couldn’t make a better WIRED pitch if you built it in a lab. Or in this case, with the help of a large language model chatbot.”How WIRED Got Rolled by an AI Freelancer | WIRED · www.wired.com →
“Margaux Blanchard didn’t exist. After an editor grew suspicious of a strange-sounding pitch and challenged Blanchard, it became clear that the work published under that name was most likely AI-generated.”Speed, hoaxes and mistrust: How AI is transforming freelance journalism | Reuters Institute for the Study of Journalism · reutersinstitute.politics.ox.ac.uk →
AI-generated victim-impact videos: disclosed recreations or unreliable courtroom evidence?
DocumentedCourt rulings say an AI-generated victim video was admitted and improperly influenced sentencing, showing a synthetic portrayal slipped into judicial safeguards and affected determinations of what was treated as real.
“finding that a judge improperly relied on an AI-generated video depicting the victim speaking about forgiveness and addressing the defendant directly.”Arizona court says judge wrongly allowed AI-generated victim video | Reuters · www.reuters.com →
“The appeals court ruled the video unfairly influenced a trial judge, and the man convicted needs to be given a new sentence.”Court rules man should be resentenced after AI video of victim was shown in court : NPR · www.npr.org →
AI-generated political content is quietly hollowing out shared trust in official election information.
DocumentedThe brief states synthetic-media weaponization is undermining trust and that people can no longer reliably distinguish AI content, directly linking slipping AI fakes to weakened public trust in official information.
“In addition, the weaponization of synthetic media has also begun to undermine people’s trust in information integrity more widely, posing concerning implications for the stability and resilience of the U.S.’s information environment.”Crossing the Deepfake Rubicon · www.csis.org →
“However, a recent experimental study by CSIS found that people are no longer able to reliably distinguish between authentic and AI-generated images, audio, and video sourced from publicly available tools.”Crossing the Deepfake Rubicon · www.csis.org →
AI hacking tools are turning software flaw discovery into a nonstop race against patch teams.
DocumentedAn official vulnerability pipeline accepted AI-made fake bug reports, and the piece warns that AI agents trusted to find and fix bugs are the ones likely to be fooled by those fakes.
“A batch of critical SQLite vulnerabilities turned out to be AI-generated fiction, yet they reached the US government's official database.”Someone posted 55 vulnerabilities in days. 54 were AI-invented. One still scored a perfect 10. · thenextweb.com →
“the systems most likely to be fooled are the AI agents now trusted to fix real bugs.”Someone posted 55 vulnerabilities in days. 54 were AI-invented. One still scored a perfect 10. · thenextweb.com →
Block released Dessa, a local speech model that talks and listens at once, with humanlike pauses and emotion.
DocumentedDessa’s voice AI is described as removing the old robotic cue and observers warn it could make scams harder to detect, indicating impersonations might bypass ordinary human‑facing authenticity checks.
“could make phone scams harder to detect.”Jack Dorsey's new voice AI is impressing users and alarming fraud watchers · www.coindesk.com →
“But with it goes one of the last easy ways to know who, or what, is on the line.”The robotic voice was a safety feature. Now it's gone · www.newsweek.com →
AI-faked research is slipping into science journals, as 1,400 papers are pulled in a day and a major journal is put on hold.
DocumentedThe story shows AI-generated papers entering journals and detection methods failing to prove misconduct, supporting that AI imitations are slipping past gatekeeping systems like peer review and automated detectors.
“AI-faked research is slipping into science journals, as 1,400 papers are pulled in a day and a major journal is put on hold.”AI-faked research is slipping into science journals… · TruthSeekers · truthseekers.com →
“Detection Text-pattern methods can identify likely AI involvement, but they cannot by themselves establish research misconduct.”AI-faked research is slipping into science journals… · TruthSeekers · truthseekers.com →
More than 400 AI apps hid abusive deepfake tools behind harmless storefronts and remote servers, slipping past platform checks.
DocumentedThe cited reporting shows ordinary-looking apps concealed sexually abusive deepfake generators and used server-side tricks to hide features, letting AI fakes bypass app-review and advertising checks.
“look like regular AI tools but had secret features that let users create violent AI-generated images and video.”Hundreds of Mundane AI Apps Are Hiding Sexually Abusive Deepfake Features | WIRED · www.wired.com →
“apps in the network contacted remote servers after installation. They transmitted information including the user's language, referral source, and a unique identifier. The server decided whether to show the app's advertised features or load a separate website with explicit deepfake generators.”Hundreds of AI apps hid abusive deepfake tools from reviewers · appleinsider.com →
OnlyFake made bypassing online identity checks as simple as ordering a custom document.
DocumentedOnlyFake’s AI-generated IDs were tested and used to pass online identity verification (KYC) checks, showing AI fakes can bypass checkpoints meant to verify authenticity.
“404 Media then used another fake ID generated by this site to successfully step through the identity verification process on OKX.”Inside the Underground Site Where ‘Neural Networks’ Churn Out Fake IDs · www.404media.co →
“the images are so good that 404 Media was able to get past the KYC measures of OKX, a cryptocurrency exchange that uses the third-party verification service Jumio to verify its customers’ documents.”People Are Using Basic AI to Bypass KYC —But Should You? - Decrypt · decrypt.co →
Every link on this board is backed by a source: 8 documented.
A man is accused of turning real minors' social-media photos into AI-generated child sexual abuse images.
9 sourcesGerman authorities did not deploy a Bundestrojaner-style spyware against AI companies’ servers.
8 sourcesMeta's Muse is quietly building dossiers on users and non-users alike, privacy experts warn.
11 sourcesUK teens may need to prove their age with ID or a face scan before entering social media.
6 sources