Briefing: YouTube gives banned channels a second chance

Our weekly Briefing is free, but you should upgrade to access all of our reporting, resources, and a monthly workshop.

Craig showed how to tell if a social media account is earning money directly from a platform.

Alexios wrote up the key findings from a working paper he co-authored with Zahra Arjmandi-Lari and Tom Stafford on the community behind X’s Community Notes fact-checking feature.

📍 You probably saw that there are tools that remove the watermark from Sora 2 videos. But did you know there’s also a website that lets you add the watermark to videos, just for the sake of chaos? (h/t Henry Ajder) Meanwhile, in a crossover absolutely no one asked for, Meta ran ads for a mobile app that says it uses Sora 2 to generate synthetic sexualized content.

📍 In more Sora 2 news: deepfake detection startup Reality Defender was able to bypass the app’s verification checks and create avatars for celebrities and CEOs.

📍 YouTube’s brief blog post about giving banned creators a “second chance,” which is likely to lead to the re-platforming of channels that ran afoul of misinformation policies, does appear to draw the line at allowing Alex Jones back on the service. The platform says creators who committed “particularly severe or persistent violations of our Community Guidelines” and those whose “on- or off-platform activity harmed or may continue to harm the YouTube community” will not be reinstated. Let’s see.

📍 Brazilian police arrested four individuals tied to a multi-million dollar Instagram scam operation that that used deepfakes of supermodel Gisele Bundchen. Also: Lupa investigated the “school of scams” community of hustle bros on video platforms, and Factchequeado released a free WhatsApp course to help users identify online scams.

📍Low-credibility websites and at least one US Congressman promoted a conspiracy theory that chemtrails have disappeared because the secret government program blasting them into the atmosphere was affected by the government shutdown (h/t NewsGuard).

📍 Authors are being targeted by AI-generated replicas, fake biographies, and workbooks (friend of Indicator Bill Adair is among them). And, well, don’t even bother searching for KPop Demon Hunters books on Amazon.

📍 Meta’s AI offered incorrect updates about the search for Gus Lamont, a missing four-year-old child in South Australia. AI was also used to create a fake image related to the boy.

📍 Eurovision News held the latest edition of its SPOTLIGHT Fact-Check & OSINT Review, where reporters talk about the tools and techniques they used to report recent stories. Reporters from Deutsche Welle, BR24 #Faktenfuchs, and Eurovision News spoke about reporting on Israeli government ads that sought to seed doubt about the extent of famine in Gaza.

📍 JSON CRACK is a free tool that displays a JSON file as an interactive graph. (via OSINT Jobs)

📍 Sketchy Boats is a free site created by Christian Panton to highlight vessels “with murky paperwork, shady ownership, strange port calls, and a suspicious tendency to vanish from tracking systems.” It assigns ships a “sketchiness score.” Panton recently spoke about the tool and their approach in a Bellingcat Stage Talk.

📍 Benjamin Strick, who works at the Centre for Information Resilience and runs a great OSINT YouTube channel, has launched a Substack. He shared some tips and insights about tracking fishing vessels, using Bing Maps for bird’s eye view, and how to create your own map of energy infrastructure.

📍 Tom Uijtdehaag wrote, “Detecting War Crimes from Space.” It’s a fascinating technical walkthrough of how the Centre for Information Resilience and BigData Republic built an open-source pipeline to detect the burning of civilian homes.

📍 SANS announced the dates for its annual OSINT Summit & Training. The one day Summit is on March 16, 2026 and virtual attendance is free if you register in advance. In-person attendance is $295 for the Summit, and extra for the six-day OSINT course that follows.

📍 Final reminder that SkopeNow’s free annual OSINT Live event is on Oct. 15. Topics include “Harnessing Overlooked Data in OSINT” and “Detecting and Investigating Deepfakes and Synthetic Media.” Craig will also present, “Using Digital Ad Libraries to Uncover Influence, Spend, and Strategy.” Register here.

📍 F-Secure’s “Scam Intelligence & Impacts Report 2025” makes a simple, alarming point: scams are getting worse all over the world. Criminals continue to innovate and to use new technology like AI to find and persuade victims.

📍 The Citizen Lab published a report about a network of more than 50 inauthentic X accounts that were “spreading narratives inciting Iranian audiences to revolt against the Islamic Republic of Iran.” It assessed that the most likely explanation is that the network is operated by “an unidentified agency of the Israeli government, or a sub-contractor working under its close supervision, is directly conducting the operation.”

📍 OpenAI’s latest threat report revealed that it banned group of Chinese-language accounts that used ChatGPT for phishing and malware campaigns, as well as a group of accounts linked to Chinese government entities that “asked our models to generate work proposals for large-scale systems designed to monitor social media conversations.” It also banned a group of Russian accounts that used ChatGPT “to generate video prompts that appeared to be intended for use with other models.”

📍Maldita and Facta co-published a two-part investigation on how AI-generated content and disinformation have helped fuel the rise of fascist nostalgia and pop aesthetics.

Want more studies on digital deception? Paid subscribers get access to our Academic Library with 55 categorized and summarized studies:

Indicator is a reader-funded publication.

Please upgrade to access to all of our content, including our how-to guides and Academic Library, and to our live monthly workshops.