An AI model trained on prison phone calls now looks for planned crimes in those calls

The model is built to detect when crimes are being “contemplated.”

A U.S. telecom company trained an AI on years of inmates’ phone and video calls and is now piloting it to scan calls, texts, and emails, aiming to predict and prevent crimes. Securus Technologies built the tool starting in 2023, including a model trained on seven years of Texas prison calls, and has been testing it in real time in jails and prisons.

The AI flags conversations for human review to detect criminal activity, including trafficking and gang-related crimes. Inmates are notified that calls are recorded, but few know their data could train AI. Advocates warn this creates coercive consent and deepens invasive surveillance, while courts have set few limits. Securus says the tool makes monitoring more efficient, not to target individuals, and aims to balance security with affordability.

🌐📲 @TechnologyBoxs