B258569
Cal. Ct. App.Jan 9, 2015Background
- Martell Chubbs was charged with a 1977 murder after DNA testing matched him to sperm on the victim’s vaginal swab; Cybergenetics’ TrueAllele software produced an extremely large likelihood-ratio match.
- Defense demanded Cybergenetics’ TrueAllele source code to test and cross-examine the software’s analysis; Cybergenetics (and Perlin) refused, asserting the code is a trade secret.
- Trial court ordered production (via out-of-state subpoena) and later compelled disclosure subject to protective measures; the People sought review in this court.
- The superior court alternatively found the confrontation clause required disclosure and indicated a protective order could be used to limit dissemination.
- The Court of Appeal reviewed whether the trial court properly compelled production of proprietary source code given trade-secret protections and applicable Discovery/Evidence Code procedures.
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| Whether trial court erred by compelling disclosure of proprietary source code over trade-secret claim | People: disclosure required for defendant to test evidence and confront expert; protective order could limit harm | Cybergenetics/Perlin: source code is a trade secret; disclosure would cause competitive harm and is unnecessary to assess reliability | Court: Yes. Trial court abused discretion; source code need not be disclosed absent a prima facie, particularized showing of relevance/necessity |
| Burden for overcoming trade-secret privilege in criminal discovery | People: trial court applied Evidence Code §§1060–1062 and confrontation concerns to require production | Chubbs: argued §§1061–1062 limit privilege-holder to protective orders/closure, not outright refusal | Court: Bridgestone burden-shifting (prima facie, particularized showing of relevance/necessity) applies; defendant failed to meet it |
| Whether TrueAllele source code is necessary to evaluate reliability of the method | Chubbs: experts say source code is the only satisfactory way to evaluate TrueAllele | Perlin: methodology, mathematical model, validation studies, and operating manuals were disclosed; source code would not meaningfully aid reliability assessment and is commercially sensitive | Court: The record shows adequate methodology disclosure and peer-reviewed validation; defense did not demonstrate how source code would materially aid their challenge |
| Whether the confrontation clause requires pretrial disclosure of privileged trade secrets | Chubbs: confrontation requires access to materials to effectively cross-examine Perlin | People/Perlin: confrontation does not trump trade-secret protection for pretrial discovery | Court: Confrontation does not authorize pretrial disclosure of privileged trade secrets; reliance on confrontation was erroneous |
Key Cases Cited
- People v. Kelly, 17 Cal.3d 24 (California 1976) (framework for admissibility of novel scientific evidence)
- Bridgestone/Firestone, Inc. v. Superior Court, 7 Cal.App.4th 1384 (Cal. Ct. App. 1992) (party seeking trade-secret discovery must make prima facie, particularized showing of relevance and necessity)
- People v. Hammon, 15 Cal.4th 1117 (Cal. 1997) (Sixth Amendment confrontation clause does not extend to authorize pretrial discovery of privileged third-party materials)
- People v. Clark, 52 Cal.4th 856 (Cal. 2011) (noting limits on invoking confrontation clause to obtain pretrial discovery)
- Commonwealth v. Foley, 38 A.3d 882 (Pa. Super. Ct. 2012) (trial court did not abuse discretion admitting TrueAllele-related testimony; source code not necessary for reliability assessment)