Often, child sex traffickers reach out to would-be clients with photos of their victims in the hotel rooms in which they're being held.
In order to help find and rescue those kidnapped children, engineers from Washington University developed an artificial intelligence system that can recognize features of a hotel room from a photo and identify where it may have been taken — potentially giving investigators new leads to fight child trafficking, according to Reuters.
So far, the app has stayed out of public hands and is used exclusively by law enforcement or groups like the National Center for Missing and Exploited Children (NCMEC) so that sensitive information like the photos of trafficking victims don't spread across the internet.
The idea is to flood the AI algorithm with hotel room photos so that it could recognize the background of a trafficking victim's photo. Some features are easy to learn, as hotel chains may standardize their furniture.
The AI system "learns a set of filters," Abby Stylianou, an engineer who worked on the project told Reuters. "Here's what a headboard looks like if it's from this hotel. Here's what a lamp looks like in that hotel."
Behind Closed Doors
Since 2016, the engineers' system has collected some 50,000 crowdsourced photos. They decided that the AI system was ready to be tested about eight months ago, according to Reuters, at which point it was shipped out to NCMEC. But so far, no one has shared whether or not the tool actually led to any successful rescues.
"We've had a couple [cases] that have given us leads and information," Staca Shehan, an executive director at NCMEC, told Reuters. "There's a lot of contributing factors that go into a recovery of a child."
READ MORE: Scientists use AI to help children sold for sex in hotels [Reuters]
More on AI forensics: AI Tool Helps Law Enforcement Find Victims of Human Trafficking