Documenting COVID-19

The Algorithms Project

The Algorithms Project, started by former Stanford JSK-HAI Fellow Garance Burke and now led by New York-based investigative journalist Georgia Gee, began with support from the John S. Knight Journalism Fellowships at Stanford University as part of the Covid Public Info project. It included contributions from Beryl Lipton and input from team members, MuckRock and Outlier Media.

The project aims to obtain information from state and federal agencies around the use of algorithms and other predictive tools amidst the COVID-19 pandemic. Specifically, the project looks into the function of algorithms in policy decisions regarding unemployment; release from state and federal prisons and jails; and surveillance, such as thermal cameras and facial recognition. These records are maintained by federal and state governments but typically are not made public without an open-records request.

Documenting COVID-19 continues to look into the use of algorithms amid the pandemic, exploring the extent of bias in AI medical technologies. The project aims to investigate whether predictive tools related to COVID-19 have had an impact on marginalized communities, such as data-driven decisions on testing locations.

If you have information or expertise that would be helpful to this project or would like to collaborate with us, you can email us at

John S. Knight Journalism FellowshipsThe Brown InstituteMUCKROCK     

Record Set(s)

North Dakota Department of Corrections & Rehabilitation Recidiviz agreement

Date Updated: September 24, 2020

Date Added: September 23, 2020


Principal Subject: North Dakota Department of Corrections & Rehabilitation

This is the Research & Confidentiality Agreement between Recidiviz, and the Idaho Department of Correction which shows how algorithms and data sharing decisions were made from January 2019 to the present.

Date Range: January 18, 2019

Format Details: .pdf (335.8KB, 6 pages)

Idaho Department of Corrections Recidiviz Data Sharing Agreement on prison release

Date Updated: September 23, 2020

Date Added: September 23, 2020


Principal Subject: Idaho Department of Corrections

This is the data sharing agreement between Recidiviz, and the Idaho Department of Correction which shows how algorithms are used for the purpose of population and release decisions in the state for the period from December 31, 2019 .

Date Range: December 31, 2019

Format Details: .pdf (939.6KB, 12 pages)

Any use of documents downloaded from this site must attribute the
"Documenting COVID-19 project at The Brown Institue for Media Innovation."

Attribution 4.0 International (CC BY 4.0)

The information on this site is available for use with attribution under a Creative Commons license, Attribution 4.0 International (CC BY 4.0), with two additional disclaimers described below. Please read through and click accept before accessing Documenting COVID-19.

For more information about required attribution, click on our Republication page. For a summary of the Creative Commons license, click here. For the full license, click here.

As this site relies on public records obtained through state and federal open-records laws, we do not guarantee the accuracy or completeness of the underlying data or documents. It may contain errors and omissions.

Public records may include personal information, which is sometimes redacted and sometimes not. However, we cannot ensure that all personal information is redacted from all materials.

If you accept both the license requirements regarding use and attribution and the two disclaimers regarding potential errors, omissions and privacy, click below to enter Documenting COVID-19.

The Brown Institute
Documenting Covid-19