Erin Kenny, Katharina Clausius, and Michael J. Crawford detail how the expanded use of DNA profile databases and genealogy in criminal investigations offers powerful tools to solve crimes and exculpate innocents but also risks permanently stigmatizing genetic relatives.
The Canadian National DNA Database was established to support forensic investigations by the Royal Canadian Mountain Police (RCMP). It now contains nearly 600,000 DNA profiles and comprises separate indices for Convicted Offenders, Crime Scenes, Victims, Voluntary Donors, Missing Persons, Relatives of Missing Persons, and Human Remains. When a crime scene sample is queried, identifications are made when DNA profiles are a perfect match: Section 6 of the DNA Identification Act (1998) prohibits the Canadian National DNA Database from implicating genetic relatives on the basis of partial matches and familial searching. As legal expert Erin Murphy puts it, such genealogical searches would “embody the very presumptions… (that) evidentiary rules have long endeavored to counteract: guilt by association, racial discrimination, propensity, and even biological determinism.” This prohibition is not unique to Canada: in the US, the FBI also forbids genealogical searches, but operating in murky legal waters, some Provincial, State, and local jurisdictions presently conduct kinship analysis.
Under the proposed Bill S-231, the Canadian National DNA Database would expand to include DNA samples from non-violent criminals and juvenile offenders specifically to facilitate genealogical searches. This database, like the FBI and other investigative agencies’ databases, “uses” CODIS, which is a system that draws upon a few dozen DNA markers to construct and search unique profiles. This analysis relies upon the integrity and purity of crime scene samples obtained: we all shed DNA, so unless (or even if) there is a discrete blood stain to probe, it can be indirectly contaminated by DNA from people who were never at the scene. One crime scene yielded DNA evidence that implicated a previously convicted felon in the rape, murder, and hiding of a victim’s corpse. The problem? He died years before the crime was committed: he had once been a plumber at the crime site and left his DNA behind.
In another case, the Phantom of Heilbronn was thought responsible for 40 or more murders that otherwise shared no common features. In fact, there was no Phantom: swabs used for sampling crime scenes were accidentally contaminated at the plant where crime kits were packaged. Contaminated analysis has also been a problem at major accredited crime labs due to sloppy cross-contamination of samples by technicians.
Contamination issues aside, partial DNA matches “have very low efficiency in locating true relatives in offender databases”: they can only expose close genetic relationships, such as parental and sibling links. When partial matches indicate a genetic relative, investigators can use Y chromosome markers or mitochondrial genome sequence data for more information. This type of analysis is also fraught with peril: investigators, judges, and jurors might not comprehend the limitations of these protocols. Michael Usry was victimized by this process when interrogated by Idaho police in 2014 about a cold-case murder-rape. Genealogy implicated Usry, and was based on Y chromosome data that his father had deposited in a consumer DNA genealogy site. It turned out that the actual killer was a seventh cousin whose only link to Usry was a common ancestor … born in 1741. Finally, Y chromosome datasets may contain profiles and data obtained through coercion from vulnerable populations such as the Roma and Uyghurs – how can justice be served exploiting such a resource?
Whereas CODIS forensic analysis is based on a few dozen DNA markers, modern consumer-oriented DNA kits reveal far more detailed information using roughly 680,000 markers. When the Golden State Killer was identified using a consumer-oriented DNA genealogy site (GEDmatch.com), law enforcement agencies immediately saw opportunities to solve serious and cold cases. GEDmatch.com was subsequently purchased by a forensics firm.
Consumer DNA kits present big ethical challenges. Service providers may not explicitly obtain consent for investigational use of consumers’ DNA profiles. Nor can it be reasonably assumed that consumers fully understand the implications such access poses for their own, or their genetic relatives’, privacy and presumption of innocence. In an interview, David Goodis, former Assistant Information and Privacy Commissioner of Ontario told us that “I can consent to the use and disclosure of my information, but nowhere else does that mean that I’m now also consenting to somebody else’s information being disclosed”. Brenda McPhail of the Canadian Civil Liberties Association takes it further: “there is absolutely no reasonable way to interpret that as informed consent from family members, up to and including unborn generations or retroactively into the past”.
Balancing public interest and privacy, investigative integrity, and the presumption of innocence is delicate. Chapter 8 of the Canadian Human Rights Act (1985) may prohibit familial searches insofar as they would represent a violation of search, seizure, and privacy provisions. Furthermore, expansion of DNA databases risks legalizing discriminatory practices. The Canadian National DNA Database is already disproportionately populated by vulnerable groups more likely to be incarcerated. Expanding the National DNA Database risks subjecting these individuals – and their genetic relatives – to unjust interrogation and suspicion. The Genetic Non-Discrimination Act (2017) provides a valuable legal framework to protect against potentially invasive, inaccurate, and discriminatory investigative practices. Our commitment to social justice demands that we scrutinize the full impact of genetic technology on the justice system.
Erin Kenny is a Student Researcher in Forensic Science at the University of Windsor.
Katharina Clausius is an Assistant Professor in the Département de littératures et de langues du monde, Université de Montréal.
Michael Crawford is a Professor of Biomedical Sciences at the University of Windsor. email@example.com