California’s Admission of AI Use Raises Concerns
CalMatters reported on Monday that California officials acknowledged last Friday the utilization of several high-risk artificial intelligence systems that hadn’t been disclosed before.
Previously, state officials had insisted that no government agency was employing high-risk automated decision-making systems, as indicated by CalMatters back in June 2025. However, the Ministry of Technology has now updated its legislative report, stating that “we currently utilize six different automated systems.” These systems can forecast, for instance, the likelihood of a recently incarcerated individual reoffending after their release.
According to the report, the California Department of Technology did not respond promptly to requests for comments.
CalMatters adds that the state is also using several systems to evaluate potential fraud in unemployment claims, administer online exams to Cal State students, and monitor cases where students may be depending on AI tools for their assignments.
One such tool, the California Static Risk Assessment (CSRA), is required by correctional regulations. This validated risk assessment tool assesses factors that are most indicative of recidivism. It assigns risk levels based on static factors like age, gender, and past criminal history. The CSRA generates numbers predicting an individual’s likelihood of being arrested for a felony within three years of parole.
According to the report, it analyzes unchanging factors about offenders to assign a risk level ranging from 1 (low risk) to 5 (high risk of violence).
The report highlights that identifying high-risk individuals allows the California Department of Corrections and Rehabilitation (CDCR) to concentrate its limited rehabilitation resources where they will be most effective for public safety.
Interestingly, the report also mentioned that CDCR staff could carry out CSRA assessments manually without using AI tools. Additionally, it pointed out six separate systems initially flagged as high risk, though they were later determined not to be.
Some critics have voiced concerns that without adequate government oversight, the public might face challenges related to transparency regarding the influence of automated systems on critical decisions, as reported by Newsweek.
