The Spit in the Courtroom: Rethinking Consent When Consumer DNA Becomes Evidence
Photo by Photo by CDC on Unsplash on Unsplash
For most of the past decade, arguments about consumer DNA have been framed as a private matter between an individual and a company. Investigative genetic genealogy has quietly rewritten that frame. When a distant cousin's cheek swab helps identify a suspect in an unsolved homicide, the person who agreed to an ancestry or health test is no longer the only person affected — and frequently has no idea that a decision was made on their behalf.
The distance between what donors believe they consented to and what their data now accomplishes is the central governance problem of forensic genetics in 2026. It is no longer a niche concern. It is a structural gap in American privacy law, and it deserves a structural answer.
A Technique That Outgrew Its Etiquette
Few methods in forensic science have spread as quickly. After the 2018 arrest of a suspect in the Golden State Killer case demonstrated that a crime-scene profile could be matched to a distant relative in a public database, agencies across the country adopted the approach. Hundreds of cases have since been closed, including decades-old homicides and long-unidentified remains.
The mechanics are unremarkable. An investigator uploads a DNA profile generated from evidence, searches for distant relatives, and then reconstructs a family tree using obituaries, census records, and newspaper archives. A conventional DNA sample from a suspect confirms the identification. Nothing in that chain requires a subpoena until the very end.
For several years the system ran on etiquette rather than law. Companies wrote terms of service. The Department of Justice issued an interim policy in 2019 and revised its guidance later. Practitioners adopted informal norms about which crimes justified the technique. Etiquette, however, is not governance, and norms do not bind an agency that never signed on to them.
Fifty Rulebooks for One Database
State legislatures have filled the vacuum unevenly. Maryland became the first state to restrict law enforcement use of consumer genetic databases, and a handful of others — Utah and Montana among them — now require a warrant or court order before a genealogical search. In much of the country, no statute addresses the practice at all. A donor in one state can be searched as a matter of routine while a nearly identical donor a few hundred miles away is protected by law.
The commercial landscape has not stood still either. The bankruptcy and subsequent sale of a major consumer testing company's assets in 2025 reminded customers that stewardship of genetic data can change hands without their participation. Terms of service are not a durable safeguard when the entity enforcing them can be sold.
Two Kinds of Consent, Not One
The most useful distinction a national standard could draw is between consent to research and consent to criminal investigation. These are not variations on a theme; they are different activities with different beneficiaries, different oversight, and different consequences for the donor.
Research participation implies aggregate analysis, institutional review, de-identification where possible, and findings that reach the public as knowledge rather than as a case number. Forensic searching implies a specific, adversarial inquiry aimed at identifying a person or that person's relatives. Bundling the two under a single checkbox is a category error, and it is the reason so many donors feel misled by searches they technically permitted.
A defensible framework would therefore be tiered and opt-in. Donors would decide separately whether their data may support academic research, identification of deceased and unidentified persons, or criminal investigations of violent offenses. Defaults should be off, changes to terms should require fresh consent, and withdrawal should be as easy as submission.
The Fourth Amendment Question Remains Unsettled
Courts have not resolved whether uploading a profile or searching a database constitutes a search for constitutional purposes. The third-party doctrine, which long held that information voluntarily shared with a company loses Fourth Amendment protection, has been narrowing since Carpenter v. United States in 2018. Forensic genealogy sits awkwardly in that gray space: the donor gave the data away, but not to the government, and not for this purpose.
A warrant requirement would not resolve the consent problem on its own. It would, however, create a record. Judges would review probable cause, agencies would have to articulate what they are looking for, and the practice would move from administrative routine into the adversarial process where it belongs.
What a Workable Standard Would Contain
A national framework need not be elaborate. Six elements would cover most of the ground.
- A warrant or court order for any genealogical search tied to a criminal investigation, with a narrow exception for identifying remains.
- Notice to the database operator and, where contact information permits, to affected donors.
- Public reporting of how often the technique is used, by which agencies, and in what categories of cases.
- Deletion rights that survive a change of corporate ownership.
- A hard separation between medical research repositories and law enforcement access.
- A clear prohibition on using the technique to investigate non-violent offenses or to identify people seeking health care, immigration assistance, or reproductive services.
The Cost of Waiting
Voluntary guidelines have accomplished more than skeptics expected, but they cannot bind the agencies that most need binding, and they cannot survive a change in leadership. The patchwork also carries a cost that rarely appears in policy discussion: erosion of trust. Consumer genetic testing and population research both depend on a public that believes participation is safe. If enough people conclude that a saliva sample is evidence waiting to be used, recruitment into legitimate studies suffers — and the studies that suffer most are those already struggling to enroll diverse participants.
Forensic genealogy is a genuine public good. It has returned names to the unidentified and answers to families who had given up. Precisely because the technique works, it should rest on something sturdier than a checkbox nobody read.