California's Static Risk Assessment — the Parole Algorithm the State Said It Didn't Have

Since being written into California's parole regulations in 2010 (15 CCR §3768.1), the California Department of Corrections and Rehabilitation (CDCR) has used the California Static Risk Assessment (CSRA) to score nearly every offender's likelihood of a felony arrest within three years of release to parole, based on static factors such as age, gender and criminal history, sorting cases into five risk tiers from Low to High Risk Violence.

CSRA is intended to inform parole decisions by classifying offenders into risk tiers using static, non-modifiable factors.

CDCR itself has noted that staff can perform the CSRA evaluation manually, meaning the automated scoring is a choice rather than an operational necessity.

A 2023 California law required state agencies to disclose their use of high-risk automated decision systems, and in 2025 the state reported using none. The Department of Technology's own June 2026 legislative report reversed that claim, revealing six systems already in active use, CSRA among them. The disclosure gap surfaced only after investigative reporting, and came the same month a bill to add guardrails on algorithmic parole decisions, SB 1248, failed in the legislature.

This is a cautionary, transparency-failure case, not a demonstrated success. A state agency reported zero high-risk automated decision systems in use, only to have that claim reversed by its own subsequent report; the fact that CSRA's automated scoring is a discretionary choice rather than a necessity sharpens the transparency question at the centre of this case.