Ben Winters

The Project

Ben investigated and published information about automated decision-making systems used in high-risk government services throughout the criminal justice cycle. Ben educated the public, advocates, and legislators working to address and combat the inherent biases in both the underlying data and algorithms used in the criminal cycle.

The increasing reliance on data and algorithms to make decisions about the length and severity of punishment among other important determinations is an underappreciated trend in the criminal justice system today. One example is the algorithm used to determine recidivism risk and to set bail, commonly referred to as a “risk assessment,” which has been shown to have disparate impacts on people of color. Other algorithms are used to determine eligibility for government benefits and more. Yet despite the increasingly significant role that these algorithms play in our justice system, they operate largely in a black box. Bringing them to light and instituting proper accountability and testing procedures will be essential to control the disparate impact these systems are having on underrepresented and over-incarcerated communities.

Fellowship Highlights

During the two-year Fellowship, Ben has:

• Testified in support of a bill establishing transparency and accountability in government procurement of automated decision-making systems in front of the Washington State Legislatures and submitted written testimony to the Massachusetts State Legislature
• Published a report called Liberty At Risk, featuring significant FOIA documents, legal analysis, and recommendations around the use of Pretrial Risk Assessments
• Worked with government agencies and other organizations to help understand and strategize about the use and impacts of automated decision-making systems.
• Published and maintained web pages highlighting open government work, legal analysis, and critical educational context

Next Steps

Ben will transition to a Counsel role at EPIC, where he will do similar work leading AI and Human Rights work both inside and out of the Criminal context.

Media