Criminal Code Reform Commission reports national sentencing analysis, recommends updates including AI 'deepfake' offense

The Criminal Code Reform Commission told the Council’s Committee on Judiciary and Public Safety on Feb. 3 that it completed two large research projects comparing District sentencing practices to national norms and that it is advancing targeted code revisions, including a proposed offense addressing nonconsensual sexually explicit imagery created with artificial intelligence.

Director Park, the commission’s executive director, briefed the committee on the agency’s recent work and submitted written materials and its FY24 annual report for the record. Park told the committee the agency’s work falls into three categories: research and data analysis to respond to critiques of the earlier comprehensive draft code (the RCCA), continuing drafting and recommendations for code revisions, and testimony and feedback on pending Council legislation.

Why it matters: The commission said its surveys and time‑served analysis show that the statutory maximums proposed in the earlier RCCA were largely consistent with national practice and that more accurate national comparisons can help anchor local sentencing policy decisions.

Park described two major research efforts. First, the commission completed a nationwide survey of statutory maximum penalties for selected serious offenses — robbery, armed robbery, carjacking, armed carjacking, burglary and armed burglary — by reviewing state criminal codes, case law and, when necessary, practitioners in other jurisdictions. Second, the commission used the Bureau of Justice Statistics’ National Corrections Reporting Program (accessed via a secure University of Michigan research environment) to build a detailed comparison of initial sentences and actual time served across states.

“We’ve now produced hundreds of graphs and charts that include detailed sentencing information from across the country,” Park said, adding the analysis included 95th‑percentile comparisons and time‑served measures that are critical because incarceration and parole practices vary across states.

Those findings, Park said, generally supported the RCCA’s approach to statutory maximums: in many states the RCCA maxima were “as long or in almost all cases longer than even the 95th percentile amount of time served for a given state,” she said.

Park told the committee the commission is also continuing to prepare and submit focused code revisions. Examples she highlighted included clarifying public‑corruption provisions (for example, misuse of nonpublic information by an official that resembles insider trading), drafting modernized animal‑cruelty provisions and recommending a new standalone offense to criminalize the nonconsensual distribution of sexually explicit images or videos produced with artificial intelligence. Park said that approach is preferable to relying on older statutes because it more precisely addresses the novel risks that AI tools create.

On the proposed AI‑created imagery offense, Park told the committee the draft includes knowledge and intent elements and narrow exceptions for clearly satirical or artful uses; she said that tailoring is meant to reduce free‑speech risk while addressing the growing harms of nonconsensual distribution: “With the rapid proliferation of more powerful and less costly artificial intelligence tools, we think that the risk of these types of misuses just continues to grow,” Park said.

Park also described the practical hurdles of producing a nationwide analysis. The commission’s research scientist, Margaret Braunschtein, led a time‑consuming process to clean and validate Bureau of Justice Statistics data stored at the University of Michigan, the director said, and the commission continues to check categorizations of state statutes to ensure apples‑to‑apples comparisons.

Park said the commission recognizes political and federal‑oversight constraints on a single, wholesale code rewrite but urged incremental, targeted revisions that can reduce litigation and confusion under the current code. She cited a recent D.C. Court of Appeals en banc reversal in Cardoso v. United States as an example of multi‑year litigation that a clearer statute could avoid; the commission’s proposal would clarify the elements that distinguish fleeting, incidental contact from kidnapping under the code.

On the commission’s near‑term agenda, Park said staff expect to deliver recommendations on anti‑dumping offenses, private‑sector corruption and an evidentiary change related to mental‑state proof that the commission believes could reduce wrongful convictions. Park said those changes are designed to be workable as standalone bills if a full code rewrite is not feasible in the near term.

The committee thanked the commission for the research and report and asked staff to provide the council with interim graphs and methodology details while the commission finishes its more detailed written report and Tableau visualizations. Park said the final methodology and cleaned comparisons will take longer because the commission is verifying state‑level categorizations and working through BJS data constraints.

No formal votes or motions occurred during the CCRC portion of the hearing.