Indicator
@indicator.media
Indicator.media publishes original reporting, in-depth investigations, and practical tutorials on OSINT tools & techniques. We provide engaged citizens and professionals with the knowledge they need to fight digital deception.
I wrote a guide to red teaming AI tools for safety for @indicator.media. It's a combination of simple frameworks and snippets of worked-through examples built from my experience at Google, running a pro bono clinic at Cornell Tech, and doing this kind of work as a journalist.
How to red team an AI tool for safety
Indicator’s no-code and product-agnostic guide to probing for harmful stuff
indicator.media
I'm quoted in today's @indicator.media newsletter, talking about the xAI lawsuit against Minnesota's deepfake nudes law: indicator.media/p/briefing-m...
Briefing: Musk sues Minnesota over synthetic nudes law
Plus: Tools for tracking foreign interference and LinkedIn launches "AI slop" button
indicator.media
Today's free Indicator Briefing is *full* of things you need to know, tools you can use, and research you should read on the digital deception beat. Check it out:
Briefing: Musk sues Minnesota over synthetic nudes law
Plus: Tools for tracking foreign interference and LinkedIn launches "AI slop" button
indicator.media
I'm preparing a guide on red teaming AI tools for safety. If you have background in this space, what are some essential resources you turn to frequently or have recommended to newcomers in this space? Here are a few I share:
New on @indicator.media: Passport bro content painting Latvia as a "leftover ladies’ paradise" where men "are more valuable than gold" has escaped containment and polluted AI results
How tabloid claims about Latvia’s “man drought” fueled a misleading social media frenzy and polluted Google results
Passport bro content objectifying Latvian women has taken over video and search platforms
indicator.media
Two Spanish operators running 20+ lookalike domains for the official sites have flooded Google with ads for {attraction name} + tickets. They then overcharge for a base ticket by a factor of 2x to 7x.
A great, detailed, guide on web scraping tools for investigations by David Weisz, who ran three different tests on seven different scrapers. His take: "I'm blown away by the progress scrapers have made over the past decade, particularly the Chrome extensions and AI tools."
No code, no budget: 7 web scrapers put to the test
Scraping no longer demands coding skills or a big budget. Here's how seven options held up across three real-world tests.
indicator.media
NEW on @indicator.media: If you're visiting the Alhambra, the Colosseum, the Louvre or one of 12 other top cultural attractions in Europe, be careful when buying tickets.
Ticket traps: Lookalike sites use Google Ads to fleece tourists at the Alhambra, Colosseum, and Louvre
Google served more than 1,500 ads for 20 sites that charge two to seven times the official ticket price
indicator.media
I'm looking for people who mistakenly booked overpriced tickets on a website that imitated the official one for any of these attractions: Alhambra, Colosseum, Louvre, Museo del Prado, Palace of Versailles, Sagrada Familia, Uffizi Galleries, Vatican Museums This is for @indicator.media
Last week, Indicator won the inaugural CJF Hinton Award for Excellence in AI Safety Reporting for our work exposing the infrastructure behind nonconsensual deepfake nude generators.
How a Haaretz reporter infiltrates companies offering disinformation as a service @indicator.media youtu.be/f6ODZiA1UEc?...
How a Haaretz reporter infiltrates companies offering disinformation as a service
YouTube video by Indicator
youtu.be
Sometimes you have to (metaphorically) kill off an emu to fact-check a shady marketing firm.
Had a lot of fun chatting to Omer Benjakob, the disinformation and cyber reporter for Haaretz, for the latest episode of @indicator.media's "Show & Tell"
How a Haaretz reporter infiltrates companies offering disinformation as a service
YouTube video by Indicator
youtube.com
WIRED and @indicator.media's tracker of deepfake sexual abuse in schools is now at more than 100 cases around the world. When we published this piece in April, the total count was under 90: www.wired.com/story/deepfa... Map and data here: indicator.media/nudifiers-un... (cc @mantzarlis.com)
New on @indicator.media: I ran an adversarial audit of Pangram, one of the buzziest AI text detectors out there. It claims a 99.98% accuracy rate. I was ultimately able to get Pangram to misidentify AI text as human 85.7% of the time in an adversarial set of 588 text samples.
What I learned running an adversarial test on an AI text detector
If you get your bot to rhyme / Pangram will misclassify it many a time
indicator.media
Excuse me but AAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHH!!!!!!!!!
We are now also on Instagram -- because who doesn't want to be reminded about scams and disinformation while looking at pics of food and faraway beaches? Follow us at: instagram.com/indicatordotmedia
Worth a read. From @craigsilverman.bsky.social indicator.media/p/i-got-insi...
I got inside a North Korean hiring scam. What I found reveals a troubling shift in tactics
DPRK hackers hired unwitting freelancers in the Philippines, Nigeria, and Colombia to put a human face on a malware operation — and make fake companies feel real.
indicator.media
You think you know your co-founder and then you find out he's interviewing to get a SWE gig for a North Korean hiring scam, smdh. Read @craigsilverman.bsky.social's awesome new investigation:
I got inside a North Korean hiring scam. What I found reveals a troubling shift in tactics
DPRK hackers hired unwitting freelancers in the Philippines, Nigeria, and Colombia to put a human face on a malware operation — and make fake companies feel real.
indicator.media
Also in depressing security news. indicator.media/p/i-got-insi...
I got inside a North Korean hiring scam. What I found reveals a troubling shift in tactics
DPRK hackers hired unwitting freelancers in the Philippines, Nigeria, and Colombia to put a human face on a malware operation — and make fake companies feel real.
indicator.media
Come i siti che indagano sui tradimenti sentimentali usano informazioni false e dubbie per farsi pagare | L'indagine di @indicator.media
Come i siti che indagano sui tradimenti sentimentali usano informazioni false e dubbie per farsi pagare | Facta
Indicator ha scoperto un ecosistema pieno di affermazioni infondate, prodotti falsi e traditori generati con l’intelligenza artificiale
facta.news
This is what always surprises people. The ability to make deepfake nudes isn't hidden in some dusty, hard to access part of the internet. It's available in the most accessible ways like the Google Play store and the Apple App store. Anyone with a cell phone can make a deepfake nude in minutes.
Alexios Mantzarlis@mantzarlis.com · 2mo agoNew on @indicator.media: We have long known that some "face swapping" apps market themselves as deepfake nude generators. At least one of these was used in a recent case in a US high school. But new research shows the majority of these tools do not have any guardrails against nonconsensual abuse.
We love to see it! Indicator is a Digital Media Awards 2026 global winner in the "Best in Countering Disinformation" category. Thanks to the World Association of News Publishers for the recognition. wan-ifra.org/events/digit...
Absolutely awesome to see that my guide to using Skills in Chrome helped an investigator with his work! Here's the post: indicator.media/p/the-indica...
joe bodnar@joeabodnar.bsky.social · 3mo agoShout out to @indicator.media for its guide on Skills in Chrome. I customized a prompt and made this great Obsidian network map in like 5 minutes.
After Indicator reached out, Apple and Google deleted 43 of the unsafe apps flagged by the audit.
New on @indicator.media: We have long known that some "face swapping" apps market themselves as deepfake nude generators. At least one of these was used in a recent case in a US high school. But new research shows the majority of these tools do not have any guardrails against nonconsensual abuse.
Most faceswap apps in the Apple and Google stores could be used to make deepfake nudes
Apple and Google removed 43 apps flagged by Indicator
indicator.media
I'm a former Community Notes super-user interviewed for this piece. Whether the notes are written by people or language-mimicking algorithms, the central flaw of Twitter's Community Notes is the same as it's always been: it's a "fact-checking" system that doesn't involve the checking of facts.
Alexios Mantzarlis@mantzarlis.com · 2mo agoToday on @indicator.media: A handful of anonymous bots have taken over fact-checking on X. This isn’t hyperbole. In the first three weeks of May, just eight AI contributors wrote 50.3% of all visible Community Notes on the platform.
Today on @indicator.media: A handful of anonymous bots have taken over fact-checking on X. This isn’t hyperbole. In the first three weeks of May, just eight AI contributors wrote 50.3% of all visible Community Notes on the platform.
8 AI bots now write 50% of X’s Community Notes
The crowdsourced moderation project faces a choice between scale and legitimacy
indicator.media
Shout out to @indicator.media for its guide on Skills in Chrome. I customized a prompt and made this great Obsidian network map in like 5 minutes.
