
Lowest Hanging Fruit: A Meta Investigation Part 2
I added 1 letter to an industry-known CSAM keyword, and Instagram delivered me the network. Part 2 of Lowest Hanging Fruit moves the investigation from Facebook to Instagram. There were no groups to infiltrate this time, so I flipped the model: I built a decoy 14-year-old around an AI-generated face pulled from the platform itself, posted with the hashtags predators search, and let them come to her. Then I mapped what followed, accounts openly soliciting child sexual abuse material (CSAM), follow graphs that work like bookmark lists, and a recommendation engine that finished the map for me under a banner reading "Suggested for You." A month after I filed my findings with the New Mexico Attorney General's office, I went back for one more look. My decoy was untouched. The WhatsApp number attached to her had taken more than 250 calls and messages in ten days. And at trial, Meta's own documents showed the company had measured from the inside what I was mapping from the outside. In its own words, from 2019: "26% of the people we recommend to Groomers were Teens." Everything in this episode was done with a laptop and a browser. It should've been more difficult. The written version of Part 2, including the evidence, is on my LinkedIn. Content notice: This series includes descriptions of predatory behavior towards children. Listener discretion is advised. Everything this series earns goes to the National Center for Missing & Exploited Children. Support the show The views expressed here are my own and do not represent those of any organization, entity, or client.
- Transcript