Courts on AI / Algorithms in criminal justice
United States v. Joyner
U.S. District Court for the District of Columbia · Federal district courts · DC · · No. Criminal No. 2023-0309 · memorandum opinion
The D.C. district court evaluates STRmix probabilistic-genotyping evidence among extensive pretrial motions in a robbery and carjacking prosecution. The opinion explains the software’s likelihood ratios and rules on admissibility limits for DNA testimony.
- Disposition
- The court granted in part and denied in part several limine motions, including challenges to STRmix DNA evidence.
What the court wrote
Verbatim from the court's document (public domain). Ellipses mark omitted text; nothing is paraphrased.
These samples were imported into a probabilistic genotyping software, STRmix, which calculated the statistical probability that Mr. Joyner’s DNA contributed to the DNA profile of the samples (“inclusion”) or did not contribute to it (“exclusion”). See Def.’s Mot. Limit DNA Evid. at 2. For each sample analyzed, STRmix produced a likelihood ratio describing “how much more likely it is to obtain the DNA results if the person of interest is a contributor to the DNA profile, rather than if an unknown, unrelated person is a contributor to the DNA profile.” Id. (citation modified). In all but one of the samples, DNA from multiple persons was detected. STRmix ultimately assigned Items 4, 44, 56(1), and 69 likelihood ratios of 10, 21, 100, and 34, respectively.
Authorities cited in the passage: Federal Rule of Evidence 702 · Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993)
compiled by the Cognesio LLP editorial teammethodas of 2026-09-07v0.1.1coding CC BY 4.0JSON Hugging Face GitHub
Hugging Face