The People v. John Wakefield
Not yet reported in New York Reports at time of opinion (previously 175 AD3d 158 (3d Dept 2019) for Appellate Division) (2022)
Premium Feature
Subscribe to Lexplug to listen to the Case Podcast.
Rule of Law:
The continuous probabilistic genotyping approach used by the TrueAllele Casework System for DNA mixture interpretation is generally accepted in the relevant scientific community under the Frye standard, and its proprietary source code is not generally discoverable for a Frye hearing or for a Sixth Amendment Confrontation Clause challenge when human analysts are subject to cross-examination.
Facts:
- On April 12, 2010, the victim was found strangled to death in his apartment with a guitar amplifier cord wrapped around his neck, and several items, including a PlayStation 3, a laptop, and a distinctive orange duffel bag, were stolen.
- Witnesses observed John Wakefield in the victim's company the weekend of the homicide, and John Wakefield admitted to three individuals that he had choked the victim.
- John Wakefield was later seen by a separate witness with a distinctive orange duffel bag (like the victim's) and attempting to trade a PlayStation and a laptop for drugs; the victim's PlayStation 3 was subsequently recovered from the home of a local drug dealer.
- Police collected DNA samples from items at the crime scene, including the front and rear outside collar of the victim’s shirt, the victim’s dorsal forearm, and a section of the amplifier cord, and also developed John Wakefield's single-source DNA profile from two bottles taken from the victim’s home.
- The New York State Police Forensic Investigation Center (Lab) initially analyzed the DNA samples, concluding that John Wakefield could not be excluded as a contributor to DNA mixtures found on the shirt collar and forearm, but the Lab's use of a stochastic threshold limited the data available for full probability calculations, especially for the complex mixture on the amplifier cord.
- The electronic raw data from the Lab's DNA testing was then sent to Cybergenetics for additional analysis using its TrueAllele Casework System, which applies a continuous probabilistic genotyping method, utilizing all generated peak data (including below the stochastic threshold), to calculate statistical likelihood ratios for John Wakefield's contribution to the mixtures.
Procedural Posture:
- Prior to trial, John Wakefield moved in Supreme Court (the trial court) to preclude TrueAllele evidence or, alternatively, for a Frye hearing, also requesting discovery of TrueAllele's software source code.
- Supreme Court granted John Wakefield's motion for a Frye hearing to determine the admissibility of TrueAllele's methodology.
- During the Frye hearing, John Wakefield's request for TrueAllele's source code was denied by the People, who did not possess it, and he made no further court motion for it.
- Following the Frye hearing, Supreme Court issued a detailed opinion finding TrueAllele generally accepted in the relevant scientific community and denied John Wakefield's motion to preclude the evidence.
- Prior to trial, John Wakefield again moved in Supreme Court for disclosure of the source code, arguing it was necessary to exercise his Sixth Amendment right to confront witnesses; the court denied this, stating the source code was not a witness and he could cross-examine Dr. Perlin.
- A jury convicted John Wakefield of murder in the first degree and robbery in the first degree.
- John Wakefield appealed his conviction to the Appellate Division, Third Department (John Wakefield as appellant, the People as respondent), which affirmed the judgment, finding TrueAllele admissible and no error in denying source code disclosure.
- A Judge of the Court of Appeals granted John Wakefield leave to appeal (John Wakefield as appellant, the People as respondent).
Premium Content
Subscribe to Lexplug to view the complete brief
You're viewing a preview with Rule of Law, Facts, and Procedural Posture
Issue:
1. Is the continuous probabilistic genotyping method employed by the TrueAllele Casework System for DNA mixture interpretation generally accepted in the relevant scientific community under the Frye standard, making the evidence admissible? 2. Does the Sixth Amendment's Confrontation Clause or discovery rules require disclosure of TrueAllele's proprietary source code to a defendant for a Frye hearing or for purposes of cross-examining an expert witness?
Opinions:
Majority - chief judge difiore
Yes, the continuous probabilistic genotyping method used by the TrueAllele Casework System is generally accepted in the relevant scientific community under the Frye standard, and no, disclosure of its source code was not required for the Frye hearing or for Sixth Amendment confrontation purposes. The Court concluded that Supreme Court did not abuse its discretion in finding TrueAllele's methodology generally accepted. The relevant scientific community recognized probabilistic genotyping as a valid DNA interpretation approach, and the foundational mathematical principles (Markov chain Monte Carlo (MCMC) and Bayes’ theorem) are widely accepted. Empirical evidence of TrueAllele's validity was demonstrated through multiple validation studies (including collaborative and independent studies, many peer-reviewed and published), its approval for forensic casework in New York by the DNA Subcommittee and the full Commission on Forensic Science, and its use by the National Institute of Standards and Technology (NIST) for standard reference materials. The Court acknowledged concerns about the proprietary nature of the technology and the developer's (Dr. Perlin's) involvement in many studies but noted that Dr. Perlin's interest was disclosed and the empirical evidence of reliability was not disputed. Disclosure of the source code was not required because it was not discoverable under the former CPL 240.20 (not a "written report or document" in the People's possession, nor was a particularized need demonstrated by motion), and arguments for its disclosure pertained more to foundational accuracy rather than general acceptance for the Frye hearing. Regarding the Confrontation Clause, the Court found TrueAllele's report testimonial but rejected the "novel argument" that the source code is the "declarant." A machine cannot be cross-examined. Instead, the human analysts who operated the system and understood its methodology (Dr. Perlin and the Lab analyst) testified and were subject to cross-examination, satisfying the right to confrontation. The defendant also failed to preserve the argument that source code disclosure was necessary to effectively cross-examine Dr. Perlin.
Concurring in result - judge rivera
No, TrueAllele's proprietary methodology was not generally accepted within the forensic scientific community without independent review of its source code, and thus, its admission was an abuse of discretion; however, the error was harmless given the overwhelming evidence of guilt. Judge Rivera argued that the trial court abused its discretion as a matter of law by admitting the TrueAllele results because the source code and underlying algorithms were kept from independent evaluators and the defense, preventing objective assessment. She stated that validation studies by interested parties (Dr. Perlin and the New York State Police FIC, who had invested in the software) are not sufficient to establish general acceptance under Frye, even with disclosure of financial interests. Endorsements from regulatory bodies like the DNA Subcommittee, NIST, and ANSI were based on incomplete information and cannot replace the courts' obligation to ensure scientific acceptance. The general mathematical principles underlying TrueAllele, while accepted, do not validate its unique, proprietary application and the inferences made by its artificial intelligence, which makes critical decisions that supplant human judgment. Without source code access, independent third parties cannot challenge these assumptions, potentially replacing human bias with algorithmic bias. For the Confrontation Clause, Judge Rivera agreed the TrueAllele report was testimonial and that Dr. Perlin was a declarant. She argued that John Wakefield was entitled to review the source code to effectively cross-examine Dr. Perlin about the software's programming assumptions and potential biases, as TrueAllele's AI provides "testimonial" statements. Despite these errors, Judge Rivera concurred in the result, finding the errors harmless beyond a reasonable doubt because the prosecution's other evidence of guilt—including unchallenged conventional DNA evidence linking John Wakefield to the crime, his undisputed presence at the scene, and the testimony of three witnesses to his confessions—was overwhelming.
Analysis:
This case significantly clarifies the application of the Frye standard to complex, proprietary scientific software in New York, particularly in the context of forensic DNA analysis. It establishes that extensive validation studies (even those involving the developer), approval by relevant scientific bodies, and use by other jurisdictions can satisfy the 'general acceptance' criterion for novel scientific methodologies like probabilistic genotyping, even without source code disclosure. The ruling also firmly rejects the notion that a computer program's source code constitutes a 'declarant' under the Sixth Amendment's Confrontation Clause, emphasizing that human experts who understand the software's methodology and parameters are the appropriate witnesses for cross-examination. This decision will likely influence future cases involving the admissibility of AI-driven forensic tools and the scope of discovery for proprietary software in criminal proceedings, making it more challenging for defendants to compel disclosure of source code solely on the basis of a general Frye challenge or Confrontation Clause claim.
