The introduction of new technology to the criminal legal space tends to follow a familiar pattern: expansive promises as to its impact, followed by an implementation that proves fraught and polarizing.
Both electronic monitoring and algorithmic risk assessments were presented as remedies to the over-reliance on incarceration; in practice, rife with unintended consequences, neither has come close to fulfilling that promise.
Given AI’s remarkable power, its use in the justice space could prove liberatory, but that would require answering some critical questions first.
Technology cannot be used to duck the hard problem of values. With the adoption of AI already underway in criminal legal systems, the time is now for humans, not machines, to ask: What do we want to use AI for?