Could artificial intelligence predict criminal behavior? A company claims its AI tool can flag signs of criminal activity before it takes place.
Telecom company Securus Technologies says it has built an AI model based on years of text, phone and video calls from prison inmates that can detect signs of imminent criminal activity. The tool has been in early use for the last year.
Speaking to MIT Technology Review, Securus president Kevin Elder said the tool can monitor communications in real time and spot "when crimes are being thought about or contemplated."
The company says models could be built for specific state or county applications, such as one that was trained solely on years of calls from inmates in Texas.
The tool is designed to screen communications for suspicious activity and alert human agents who then look more closely at the content.
Securus says it has helped stop human trafficking, smuggling and gang activity, but it did not give specific examples of how AI may have been used in the process and has not said where its AI tools are deployed now.
RELATED STORY | Lawmakers press tech and health experts on AI safety and data privacy, hoping to shape future regulation
While prisoners' calls and other communications are often recorded, rights advocates say inmates have not consented to training AI on the data.
Recent regulation has also shifted some costs of AI development onto inmates. In June of 2025, FCC Commissioner Brendan Carr wrote that the agency was pushing back deadlines that would have set new caps on the rates charged for prison communications — and allowing more use of AI tools in the space.
"With today’s actions, we’re helping to ensure that communications are more readily available and that important safety and security protocols are maintained. This includes steps that can lead to broader adoption of beneficial public safety tools that include advanced AI and machine learning," Carr wrote.
