Diana
I was born in Timisoara Romania but raised in Kitchener, ON, Canada all my life. While I was recently working for an AI company, I began a collaboration with a Marine Corps veteran and former intelligence officer who was working on a very important project I took an immediate interest in. He had acquired a research institute which had been charged with helping the U.S. Department of Homeland Security (DHS) identify, locate and in some cases even help rescue some of the over 300,000 undocumented immigrant children that the U.S. government had somehow lost contact with between 2021 and 2024. What started out as merely carelessness and callous negligence within the Health and Human Services bureaucracy of the U.S. quickly spiralled out of control into one of the largest humanitarian crisis in North America in a generation. As of summer, 2025, the U.S. government had recovered fewer than 20,000 of these missing children, and approximately 33,000 had been discovered to be completely gone. Sponsor addresses where the children were delivered to were either non-existent, or in many cases they were a vacant lot or an open field, meaning the person picking up the child deliberately meant for them to disappear. No one in the government can yet explain how such a bizarre and horrifying nightmare could have unfolded, but internal investigations (and potential criminal inquiries) have slowly begun.
As the public became aware of this catastrophe, my friend began investigating ways to leverage AI to analyze the various government databases — something akin to the DOGE framework. Together, we refined the vision: to deploy advanced machine learning models to help re-examine and closely data-mine all information relevant to the times, routes, transportation, locations, affiliated entities and sponsors for the 300,000 unaccompanied migrant children who disappeared in the U.S. between 2021 and 2024. Certain aspects of the data mining are quite simple, but other components are extremely complex, sensitive and expensive. Consequently, any private effort that has a chance of being effective needs well known advocates and public figures to help disseminate the message to the public and engage in complex strategic communications and outreach. And yet this humanitarian mission is still gaining national momentum. My Marine colleague has secured a meeting in Washington D.C. this fall to present the project’s proposed scope of work, task organization for staffing, and other relevant findings along with the technological roadmap that his data miners will need assistance with. As one of the initiative’s development partners — and a recognized public figure in the modelling world — I’ve been invited to serve as a digital ambassador and advocate for the campaign. I wish more than anything only to lend my small voice to the 300,000 souls who have had their voices silenced. While some may in fact be OK with some benevolent stranger who has taken them in, so many are confirmed to have been trafficked into horrible cases of abuse, neglect, wage slavery, sex slavery, or possibly worse. I am honored to use my voice to help amplify the mission and catalyze action toward recovering these vulnerable children.
-Diana Serban
