Craig Silverman

I left my job at ProPublica about a month ago. Here's why: I've teamed up with @mantzarlis.com to launch indicator.media, a new publication dedicated to exposing digital deception and to equipping professionals and the public with knowledge and skills to investigate it. Sign up for free below!

New on @indicator.media: There's a whole lot of online manipulation on Wikipedia that's just sitting there ready to be explored. With much help from journalists, researchers, and Wikipedians, I put together a guide on how to dig into the data.

Today on @indicator.media: Google is breaking its promise to label ads for unofficial government services providers. Travelers are paying the price.

Excited to be part of this!

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.

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...

In May's Show & Tell, @craigsilverman.bsky.social and I are joined by Amnesty International's Barbara Marcolini. Barbara spoke about her role as a journalist in a human rights org. and walked us through how she documented the Israeli military's destruction of civilian infrastructure in Lebanon.

Today on @indicator.media: Would you be willing to live alone in a remote Norwegian cabin for 90 days in exchange for $8 million? What about forgoing your phone to stay in a tropical overwater villa for $3.7 million?

Today on @indicator.media’s free Briefing: Google ran ads on search results for queries about removing nonconsensual intimate imagery (NCII). Most of the ads promoted generic content removal products that don’t provide a tailored service for NCII. Most importantly, all of them are paid services.

New on @indicator.media: With @mattburgess1.bsky.social at @wired.com, I built a map of the reported cases where AI was used to create nonconsensual intimate imagery within schools around the globe.

Was fun to contribute to this great, free training initiative from GIJN!

We did a great podcast with @koltai.bsky.social where she walked through the tools and techniques used to expose the infrastructure and people behind abusive AI nudifier websites. It's a fun OSINT tutorial about holding awful people accountable: www.youtube.com/watch?v=vn58...

On the pod, @craigsilverman.bsky.social, Craig’s cat Rosie, and I walked thru these investigations with Kolina. Kolina showed us how she uses tools like DNSlytics, DomainTools, and OpenCorporates to connect nefarious websites together and reveal their ultimate owner(s). Check it out!

Great to talk to @craigsilverman.bsky.social from @indicator.media about why 2026 is a year of change in the beneficial ownership world - and what it means for people trying to find out the real owners of companies and trusts indicator.media/p/finding-th... #beneficialownership #OSINT #GODIN

I bet you a loonie that Musk’s embrace of "unreliable information" -- actual false news – "extreme hyperpartisan rhetoric," "AI slop, & sexualized imagery" will both undermine AND derail Bier’s product roadmap, @craigsilverman.bsky.social: it's a structural conflict. indicator.media/p/briefing-x...

Today's free Briefing looks at how Musk's embrace of extreme politics and slop threatens to derail X's product roadmap. We also cover essential news and tools for digital investigators: Telegram search, a geolocation model you can run locally, and more. indicator.media/p/briefing-x...

Wrote up a little guide to building your own reverse image search engine for @indicator.media, steered by the brilliant fellow @towcenter.bsky.social alum @craigsilverman.bsky.social: indicator.media/p/the-indica...

Cool guide out today on @indicator.media! Reverse image search is a valuable verification tool. But sometimes you want to search within an image bank that is not public. @ishaan-jhavs.bsky.social explains how to do it:

NEW on @indicator.media: Tech companies have promised to help limit the damage to online truth-seeking by labeling AI content. But Indicator's latest audit found that this labeling infrastructure has multiple failure points

A real doctor hid behind a pseudonym to flood social media with AI-generated health slop — and sell books off the back of it. He accumulated tens of millions of views & millions of followers and sold books like "Reverse Type 2 Diabetes Naturally.” New investigation: indicator.media/p/a-real-doc...

Folks, this is best edition of our free newsletter that we’ve ever published. Absolutely jam packed with insights, analysis, and tools to make you a smarter consumer and investigator of digital information. Check it out:

Check out the latest episode of @indicator.media Show & Tell! Maldita.es has been doing some incredible work investigating scams, disinformation, and worse on the biggest tech platforms. So Craig and I got @cjimenezcruz.bsky.social to walk us through three of them, spreadsheets in hand.

Toronto's @thestar.com sent a cease and desist letter to the creator of the podcast network that @indicator.media wrote about on Wednesday. He has promised to reverse course. www.thestar.com/opinion/publ...

Here's a video about my recent investigation into more than two dozen accounts using AI-generated doctors and health professionals to push dubious advice and sell products on Instagram, Facebook, TikTok, and YouTube: indicator.media/p/ai-generat...

NEW on @indicator.media: I found an AI-generated podcast network that has published 350,000 episodes in the past month. That's 11,000 episodes a day, or about one year's worth of audio content. You'll be shocked to find out it's not generating original material.

Ecommerce fraud is rampant. It's also easy to launch and hard to trace. My new guide to investigating ecommerce sites covers how to ID the store platform, trace ownership, analyze traffic, do a safe test purchase etc. I also found a few free, new-to-me tools: indicator.media/p/the-indica...

Last week I speculated that by the end of the year a majority of “helpful” Community Notes on X could be AI-generated. To monitor this sorry milestone (and rate my forecasting abilities), I decided to build a free dashboard that tracks the share of AI notes. Check it out on @indicator.media

New on @indicator.media: I speak to San Francisco City Attorney David Chiu about his office's landmark lawsuit against 22 AI nudifiers.

Today on Indicator: How much does a TikTok account cost? Well, it depends. But on one prolific marketplace I track, you can get a handle for anywhere from 95¢ to $235.

Craig and I were tired of going to the Bar for Newsletter Dudes and getting asked "Bros, do you even *have* a podcast?" -- so we launched one. Our first guest is @jeffhorwitz.bsky.social of "Facebook Files" fame. Jeff loves LinkedIn and handling bees. Check out Show & Tell from @indicator.media:

New on @indicator.media: "@grok is this true" was the single most frequent reply tagging X's AI chatbot in the six months following its launch.

Synthetic doctors are outperforming real ones on social media. I found more than 25 AI health accounts with +8.5 million followers. The Canadian Medical Assoc runs a program featuring real doctors. Its prez told me: "I would say that our following is not as large." indicator.media/p/ai-generat...

Apple taps Google to power Siri as @leolaporte.me, @jeffjarvis.bsky.social, @paris.nyc, & @craigsilverman.bsky.social discuss tricky AI scams on Intelligent Machines, covering Meta refocusing on AI hardware and Senate's nonconsensual deepfake bill! twit.tv/shows/intelligent-machines/episodes/853

Just a heads up that this is the best time to join @indicator.media as a member because it's 20% off. We'll keep you on top of OSINT tools/trends, build your investigative skills, and deliver exclusive reporting about digital deception/disinfo. Plus monthly workshops! Info: indicator.media/upgrade

Glad @americansunlight.org's research with @indicator.media has been cited in @theguardian.com, along w my comment on the infrastructure of deepfakes. I've included some of the rest of my comment below - "if a man can’t undress me without my consent on the street, he can’t do it online either." 🙅🏻‍♀️

New: I found more than 20 deepfake videos re-enacting or re-imagining the final moments of Renée Good's life. These videos mocking a tragic death, which I'm calling "digital desecrations," are the most extreme version of a pervasive trend within generative AI of 'resurrecting' the dead.

Indicator has covered this subject extensively. Here's all of our reporting on AI nudifiers, platform failures, and countermeasures in one place: indicator.media/nudifiers-un...

It's been exactly one year since Mark Zuckerberg pulled the plug on the US arm of Meta's fact-checking program. But the program trundles on everywhere else in the world: I was able to confirm the company has renewed many contracts in 2026, albeit at a steep discount.

One example of how Grok is being used to target women. Swedish Deputy Prime Minister Ebba Busch being sexualised, degraded, and humiliated step-by-step by Grok. All the images accurately reflect the prompts provided.

This annual survey from the Canadian Medical Association has important (and often sobering) data about the state of our information landscape. Well worth attending:

America in 2025: Disinformation researchers get their visas revoked, while two DOJ-investigated Russian information operation assets are welcomed back for Christmas.