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

MIT
MIT
1w ago • 11 views
Securus Technologies is testing an AI model to analyze inmate communications for crime prevention. This raises concerns about privacy and consent.
An AI model trained on prison phone calls now looks for planned crimes in those calls
A What happened
Securus Technologies is piloting an AI model designed to analyze communications from inmates, including phone calls, texts, and emails, to predict and prevent crimes. The company has trained its AI using a vast database of recorded conversations, claiming it can detect criminal activity early. While Securus asserts that its tools have disrupted serious crimes, critics argue that inmates are not adequately informed about how their data is used. Concerns about privacy and consent are heightened by the fact that inmates pay for these calls. Recent regulatory changes allow Securus to fund its AI initiatives using fees collected from inmates, which has sparked backlash from prison advocates. The Federal Communications Commission has been involved in discussions about the funding of these monitoring tools, with some officials dissenting against the financial burden placed on incarcerated individuals and their families.

Key insights

  • 1

    AI for Crime Prevention

    Securus uses AI to analyze inmate communications for early crime detection.

  • 2

    Privacy Concerns

    Critics argue inmates are unaware their data trains AI models.

  • 3

    Regulatory Changes

    Recent FCC reforms allow telecoms to charge inmates for security costs.

Takeaways

The use of AI by Securus Technologies to monitor inmate communications raises significant ethical and privacy concerns, particularly regarding consent and the financial implications for incarcerated individuals.

Topics

Technology & Innovation AI & ML World & Politics Policy & Regulation

Read the full article on MIT