Hundreds of people incarcerated in Arizona have failed to update critical prison management software longer than they should be, according to a whistleblower report.
Arizona Department of Corrections employees have known about the bug since 2019, Phoenix-based NPR affiliate KJZZ reported. The flaw follows a change in state law that the software simply cannot handle and has not been updated to deal with it.
Arizona has one of the highest jail rates in the country, with drug possession convictions a major factor behind the numbers. A June 2019 Arizona State Law Amendment (PDF) created a mechanism by which inmates convicted of certain non-violent drug offenses can earn credits toward early release. Eligible inmates who complete a program such as a GED equivalent or drug treatment while incarcerated can earn three days credit for each seven days served and reduce their time behind bars to 70 percent of the time. sentence assigned to them.
The Arizona Department of Corrections FAQ (PDF) on the program states that inmates and their families should not contact anyone to request an eligibility review. “This is done automatically based on the system’s programming, which generates a list for Time Comp,” the document reads. “Reviews are being completed based on the earliest scheduled release dates.” According to whistleblowers, however, this system cannot identify eligible participants and still does not do these calculations.
“We knew from day one that it wasn’t going to work,” a Department of Corrections source told KJZZ. “When they approved this bill, we looked at it and we said ‘Oh shit.'”
Whistleblowers said they began sending “repeated internal warnings” to the department’s IT officials in 2019. KJZZ obtained a copy of an October 2020 bug report detailing how the software, called ACIS, is breaking the law .
“Currently, this calculation is not in ACIS at all,” the report reads. “ACIS can calculate 1 credit earned for every 6 days served, but this is a New calculation.”
The ministry confirmed to KJZZ that it was aware of the problem with ACIS. “The data is calculated manually and then entered into the system,” spokesman Bill Lamoreaux told KJZZ, saying the department has identified at least 733 inmates who are eligible to participate in the early release program but who are not yet registered.
Prison (poor) management software
There is a whole sprawling “patch” software industry in the United States. At one level, prison and inmate management systems make sense: you need to know who is in an institution, where they are, what their medical needs are, who their legal representation is, etc.
However, as parts of the criminal justice system are more and more entrusted to management algorithms, real people are more and more affected by the flaws in these software packages. The deployment of ACIS in Arizona took three years longer than expected and consistently went over budget, local media reported at the time.
Since it was finally implemented over Thanksgiving weekend in 2019, ACIS has remained plagued by bugs, sources told KJZZ. Several modules of the software did not function properly, “including modules that track inmate health care, enumeration, inmate assets, commissary and financial accounts, religious affiliation, classification of security and gang affiliations “. And when human error introduces something incorrect into an inmate’s file, employees may not have a way to remove or correct the error.
“In one case, a disciplinary measure was mistakenly entered in an inmate’s file,” a source told KJZZ. “But there’s no way to go back. So this guy was punished and he couldn’t make a phone call for 30 days. That’s the kind of thing that gnaws at you every day. “
Software suites used by law enforcement and courts – long before anyone was subjected to prison management software – have also been widely criticized in recent years, both for their lack of transparency and for their prejudices. racial. ProPublica published an in-depth and detailed report in 2016 finding that the courts’ risk assessment software ranked black suspects as more likely to commit other crimes than white suspects who allegedly committed similar acts.
Researchers in 2018 analyzed the software and found that groups of random amateurs had the same accuracy when projecting suspects’ recidivism rates; people and algorithms got it wrong about a third of the time.