Our weekly Briefing is free, but you should upgrade to access all of our reporting, resources, and a monthly workshop.
This week on Indicator
Craig published the Indicator Framework, a repeatable method for evaluating the strengths and weaknesses of digital assets like an IP address, email, crypto wallet, and more. It’s free to read, but members get access to a Matrix that enables you to visualize your evidence, and to a dataset of more than 70 indicators commonly gathered in OSINT. He’ll walk through a case study of applying the framework in today’s members-only workshop at 11 ET.
Alexios wrote about the "passport bro" videos selling Latvia as a "leftover ladies' paradise." He tracked 89 viral posts (some with up to 12 million views) back to a fake "husbands by the hour" tabloid story, a quote recycled from a 16-year-old BBC article, and viral photos of women who are actually in Poland.
Also, Alexios was quoted in a Reuters Institute article about AI text detection and journalism.
Deception in the News
How to check a Substack post with Pangram. Image via Substack
Substack launched a “Scan for AI text” feature, powered by Pangram. CEO Chris Best made the perfectly sensible argument that “when content made by no one takes over parts of the internet that are supposed to be human, it pollutes the commons and makes it hard to discover and hear human voices.” He also floated some additional product tweaks, including the option to see less AI in your feed.
I moved my newsletter off Substack in February 2025 and think its leadership is prone to presenting economically beneficial decisions as principled stands. But besides a sideswipe at LinkedIn (“we don’t want to wait until your Substack app turns into LinkedIn before we start to learn and make progress.”), I don’t see much in this decision that is necessarily self-serving. Am I missing something? — Alexios
📍 Meta used a novel legal argument to secure dismissals of two class-action lawsuits over "pump-and-dump" stock scams that were advertised on its platforms. The company argued the suits were securities fraud claims in disguise and that a federal law barring exactly that kind of claim applied. As Reuters columnist Jenna Greene noted, "The closer the alleged fraud is tied to a stock purchase, the harder it may be to pursue a consumer class action against the company."
📍 The American advocacy group, “Run for Something,” reportedly trained Democratic candidates to post positive content about themselves on Reddit in the hope that it might be picked up by chatbots.
📍 An Arkansas photographer was charged with using AI to turn photos of his underage clients into child sexual abuse material.
📍 Late last week, San Francisco City Attorney David Chiu sent letters to Apple and Google demanding the removal of 13 faceswapping apps that could be used to create nonconsensual deepfake nudes. (Alexios interviewed Chiu earlier this year and wrote about the lack of guardrails on faceswapping apps last month.)
📍 Fact-checkers Aos Fatos and Agência Lupa were among the Brazilian publications that forswore any advertising, sponsorship, or other commercial partnerships with online betting platforms.
📍 The Trump administration is working on a UN resolution condemning government interventions against online misinformation as an attack on free expression.
📍 AI slop is threatening the value of citizen science in birdwatching.
Tools & Tips
📍 Archivarix Echo is a search engine that pulls results from more than 200 online archives, including the Wayback Machine, national archives, academic research databases, and more. It also has Tube Search, which can “Find deleted YouTube videos and recover their metadata, thumbnails, and subtitles from archive data.” (Via Cyber Detective)
📍 Yehor Selin launched Shadow Simulator #01, a free OSINT course where you play the role of an investigator on a compliance desk. Available in Ukrainian and English. More info here.
📍 Ritu Gill of Forensic OSINT published a free OPSEC Field Guide.
📍 Matthew Facciani published the Lateral Reading Companion, a free tool that “walks you through checking whether something is true, the way professional fact-checkers actually do it.”
📍 Toddington International published “Building Your First AI Stack: Start Small, Think Strategically.”
📍 Fed Gov Today held a series of conversations around the topic, “From Collection to Cognition: How the OSINT Community Is Navigating the Age of AI and Data Overload.” You can watch the videos here.
📍 Mike Reilley of JournalistsToolbox.ai published “Pangram and Tips for Detecting AI in Writing.”
📍 The Global Investigative Journalism Network published “Reporter’s Guide to Investigating Cryptocurrency.”
📍 Arab Reporters for Investigative Journalism recently held a webinar, “Open Source Intelligence (OSINT) & Cross-Border Investigative Collaboration.” You can watch the presentation on YouTube:
Events & Learning
📍 ESPY is giving a free webinar on July 29, “Arabic Monitoring for OSINT Investigations.” Grab the link here.
Reports & Research
📍 The Israeli government funded an influence operation targeting Americans with millions of AI-generated texts, according to The Wall Street Journal.
📍 An analysis by LetsData, a threat detection platform, found that “AI-generated personas—including fake soldiers, veterans, recruiters, and a rabbi—are flooding Ukrainian-language social media.” It said the synthetic personas amassed more than 20 million views in the first half of this year.
📍 9to5Mac found more than 60 apps in Apple’s Brazilian App Store posed as gaming apps but were actually a gateway to gambling services.
📍 The New York Times published a video about “hundreds of A.I.-generated doctors, healers and wellness influencers on social media touting the health benefits of supplements to American consumers.” (Related: PolitiFact launched Well Informed, a new service focused on debunking health misinformation.)
Want more studies on digital deception? Paid subscribers get access to our Academic Library with 75 categorized and summarized studies:
One More Thing
ESPN, the American cable sports giant, recently had a round of layoffs. That led some X users to express hope that their least favorite on air talent might be also be let go. Jemele Hill left ESPN almost a decade ago, but one anon X account tweeted that she “should be next.”
Hill responded with a joke tweet that said, “I really hope ESPN lays me off … In fact, let me just submit my resignation to make it easier on them.”
That’s when the content farms took over. Her sarcastic tweet led to real articles claiming that she had resigned from ESPN. Hill said it shows what’s wrong with the current information ecosystem and “why telling people to Google information doesn’t work.”
Indicator is a reader-funded publication.
Please upgrade to access all of our content, including our how-to guides and Academic Library, and to our live monthly workshops.
