New York Assembly Bill A11320

New York 2025-2026 Regular Session

• Introduced: 5/11/26
• Refer: 5/11/26

Caption

Prohibits the department of corrections and community supervision from using artificial intelligence in parole determinations

Summary

This bill would amend New York’s executive law and correction law to bar the Department of Corrections and Community Supervision from using artificial intelligence in two related areas: parole determinations and transitional accountability plans. In the parole context, the department could no longer use AI, AI models, or AI systems to evaluate risk and needs factors, assess rehabilitation, decide whether an incarcerated person should be released to parole supervision, or determine the level of supervision after release. The bill also requires that each release decision and supervision level be approved in writing by the Board of Parole or its designee. The bill also changes the law governing transitional accountability plans, which are individualized case management plans developed when an incarcerated person enters custody. Those plans are intended to guide programming, treatment, and reentry preparation during incarceration and community supervision. Under the bill, the department would be prohibited from using AI in developing those plans, and each plan would need written approval by the commissioner or the commissioner’s designee. The bill takes effect immediately.

Impact

If enacted, the bill would add explicit statutory limits on the use of automated decision-making tools in correctional and parole administration. It would amend Executive Law section 259-c and Correction Law section 71-a to require human approval for parole release, supervision levels, and transitional accountability plans, while preventing the department from relying on AI technologies defined under the General Business Law and State Technology Law. The practical effect would be to preserve individualized, human-reviewed decision-making in parole and reentry planning and to constrain the department’s ability to use algorithmic risk assessment or generative AI tools in these processes.

Sentiment

The available context suggests a generally cautionary or skeptical stance toward AI use in correctional decision-making, with the bill framed as a safeguard against automated influence over parole and reentry planning. Because there are no committee transcripts or recorded votes provided, there is no direct evidence of support or opposition from legislators in the available record. The bill’s caption and text indicate a policy preference for human oversight and accountability rather than technology-driven determinations.

Contention

The main point of contention is likely whether artificial intelligence should have any role in parole and correctional case management. Supporters would view the bill as protecting fairness, transparency, and individualized judgment in high-stakes decisions affecting incarcerated people’s liberty and supervision conditions. Opponents, if any, would likely argue that AI tools can assist with consistency, efficiency, and risk assessment, and that a categorical ban could limit administrative flexibility. Another possible issue is the breadth of the prohibition, which covers not only parole decisions but also the development of transitional accountability plans and requires written approval at multiple stages.