by David Kim

Introduction: A Booking Procedure That Does Not End at Booking

In late January 2026, a Minneapolis resident identified only as Ben drove with his wife to a street where federal immigration officers had been spotted. He stood on the sidewalk and filmed. According to an NPR investigation published in March 2026, a masked officer ran at Ben and tackled him to the ground. He was handcuffed, taken into custody, and subjected to what has become a routine part of the federal arrest process: a buccal swab of the inside of his cheek, collecting a sample of his DNA. Ben was released. He was not convicted of anything. NPR reported that in the past, DNA specimens taken by federal immigration officers have been added to a national database maintained by the FBI, but it also reported that it was unclear where the samples acquired from protesters in recent months were ending up or how they were being used.

NPR found five other people in Illinois, Oregon, and Minnesota who described similar experiences in recent months, each arrested during protests against Immigration and Customs Enforcement (“ICE”) operations. In sworn statements fileßd as part of lawsuits challenging the administration’s enforcement tactics, they reported that officers took or attempted to take DNA samples after what the individuals and their attorneys contend were constitutionally protected acts of political expression. The Department of Homeland Security responded that federal law requires DNA collection from anyone who is arrested or facing charges.

The practice is not new. In March 2009, Lily Haskell attended a peace rally at San Francisco’s Civic Center to protest the Iraq War. When she allegedly tried to assist a fellow protester who had been taken into custody, police arrested her on suspicion of a felony. At the city jail, Haskell was told she had to provide a DNA sample or face an additional misdemeanor charge. She was also warned that requesting an attorney before complying would result in her being held until arraignment. She submitted. No charges were ever filed against her. But Haskell’s genetic information entered the California state database and, from there, the national system accessible to law enforcement at every level. More than a decade and a half later, the ACLU of Northern California reported that the sample and profile had still not been removed.

These are not isolated cases. They are products of a legal and institutional architecture that has grown steadily since the early 2000s. As of this writing, 31 states, the District of Columbia, and the federal government authorize the collection of DNA from people who have been arrested for, but not convicted of, designated offenses. (The National Conference of State Legislatures has counted 34 states, but that figure includes two states whose arrestee provisions have been struck down by state courts and one whose statute functions as conviction-only in practice.) The FBI’s National DNA Index System held more than 6.1 million arrestee profiles as of November 2025, alongside more than 19 million offender profiles. The system spans local, state, and federal databases linked through the FBI’s Combined DNA Index System, known as CODIS, which allows profiles to be searched across jurisdictions by hundreds of participating laboratories nationwide.

The legal foundation for all of this rests on a single United States Supreme Court decision. In Maryland v. King, 569 U.S. 435 (2013), the Court held by a five-to-four vote that collecting a cheek swab from a person arrested for a serious offense is a reasonable booking procedure under the Fourth Amendment, comparable to fingerprinting or photographing. Justice Anthony Kennedy, writing for the majority, framed DNA collection as an identification tool that helps officers determine who they have in custody and whether that person poses a risk. Justice Antonin Scalia, in a dissent joined by three colleagues, warned that the decision could not be limited to its stated rationale. The actual function of the DNA search in King, Scalia observed, was to investigate whether the arrestee had committed other crimes, a purpose the Fourth Amendment does not permit without individualized suspicion.

More than a decade later, the trajectory of DNA collection has vindicated the dissent’s warning. What was publicly justified as a narrow identification measure at the booking window has evolved into a far larger enterprise. Legislatures have expanded the categories of offenses that trigger collection. In 2026, Missouri advanced legislation to expand DNA collection from arrests for first- or second-degree burglary and felony offenses under Chapters 565, 566, 567, 568, and 573 to all felony arrests, and Iowa moved similar legislation through its House. Federal agencies now collect DNA from immigration detainees and, according to recent reporting, from protesters engaged in political speech. The biological sample itself, which contains a person’s full genetic blueprint, is frequently retained alongside the digital profile. And the expungement mechanisms that were supposed to protect the unconvicted often function poorly in practice, requiring individuals to navigate bureaucratic processes that many cannot access.

The result is a system that operates on a logic far broader than the one courts and legislators invoked to justify it. A person who is arrested, held for a few hours, and released without charges can find their genetic information stored permanently in a searchable law-enforcement database. That profile can be compared against evidence from unsolved crimes across the country. It can be used to identify the person’s biological relatives. It remains available even when the legal basis for the original arrest disappears. What was sold to the public as a modest administrative step at the front end of the booking process has become, on the back end, a durable instrument of genetic surveillance affecting millions of people who have never been found guilty of a crime.

This article examines how that transformation happened. It traces the legal, legislative, and institutional machinery that turned a booking-room swab into permanent inclusion in a national law-enforcement database. It explains what DNA collection actually involves, why it is qualitatively different from fingerprinting, and how courts came to accept the analogy that made the expansion possible. It presents the government’s strongest justifications fairly, then tests them against the constitutional, scientific, and structural evidence. And it asks the question that the current framework has largely avoided: when the state claims the power to collect, store, search, and share a person’s genetic identity before that person has been convicted of anything, what remains of the presumption of innocence?

How the System Actually Works

Before examining the constitutional questions that surround pre-conviction DNA collection, it helps to understand what the process actually involves. The public debate tends to collapse everything into a single image: an officer swabbing the inside of someone’s cheek. In practice, what follows that swab is a chain of scientific procedures, database entries, cross-jurisdictional searches, and retention decisions that extends far beyond the booking room. Each link in that chain matters, because each one expands the state’s informational power over a person who may never be convicted of a crime.

Sample vs. Profile

The first distinction that gets lost in public debate is the difference between a biological sample and a forensic profile. When an officer runs a cotton swab along the inside of an arrestee’s cheek, the swab collects epithelial cells containing the person’s complete DNA. That physical sample is a biological specimen, and it carries the full breadth of an individual’s genetic information: ancestry, heritable health conditions, predispositions, and family relationships. It is stored by the collecting laboratory, often on a card or in a sealed container, and in many jurisdictions it can be retained indefinitely.

The forensic profile is something different. A laboratory extracts DNA from the sample, then analyzes a set of specific genetic markers known as short tandem repeats, or STRs. Since January 2017, the FBI has required analysis of 20 designated STR loci for profiles entered into the national system. These loci are regions of non-coding DNA, meaning they do not directly reveal traits or medical information. The resulting profile is essentially a string of numbers representing the alleles found at each locus. According to the National Institute of Justice (“NIJ”), it is this numerical profile, not the underlying biological material, that is uploaded into law-enforcement databases and used for comparison.

Supporters of DNA collection often emphasize the profile’s limited informational content. The FBI’s own fact sheet stresses that CODIS stores no names or personal identifiers alongside the numerical data. But this framing obscures a critical point. The biological sample, from which far more sensitive information can be extracted, is typically retained by the laboratory even after the profile has been generated. As long as the sample exists, the possibility of re-analysis using more advanced techniques remains open. The profile may be a set of numbers. The sample is a person’s genetic blueprint.

CODIS, State Systems, Local Databases, and Federal Sharing

DNA profiles enter a tiered database system that connects local law-enforcement agencies to a national network. At the base sit Local DNA Index Systems, maintained by individual crime laboratories. Above them are State DNA Index Systems, which allow laboratories within a state to share and compare profiles. At the top sits the National DNA Index System, or NDIS, administered by the FBI. The software platform that links all three levels is the Combined DNA Index System, known as CODIS.

As of November 2025, NDIS contained more than 19.2 million offender profiles, more than 6.1 million arrestee profiles, and approximately 1.4 million forensic profiles derived from crime-scene evidence. Over two hundred accredited laboratories participate in the system. When a crime-scene profile is uploaded, CODIS automatically searches it against the stored offender and arrestee profiles at all three levels. A match, known in the system’s terminology as a “hit,” generates an investigative lead that is reported back to the submitting agency. As of November 2025, CODIS had produced more than 781,000 hits and assisted in more than 758,000 investigations.

What sounds local at the moment of collection quickly becomes national. A profile generated by a county sheriff’s office in Iowa can be searched against forensic evidence from an unsolved case in Florida, a cold-case sample stored in New York, or an evidentiary profile submitted by a federal agency. State participation in NDIS is governed by memoranda of understanding with the FBI Laboratory, which set operational standards for uploading, searching, and expungement. Federal law permits disclosure of stored DNA samples and analyses to criminal justice agencies for law-enforcement identification purposes, in judicial proceedings if otherwise admissible, to criminal defendants for defense purposes in their own cases, and – if personally identifiable information is removed – for population statistics, identification research and protocol development, or quality control. Unauthorized disclosure of individually identifiable DNA information is punishable by a fine of up to $100,000, and unauthorized obtaining of DNA samples or individually identifiable DNA information can carry a fine of up to $250,000 and up to one year in prison. But within that authorized network, the reach of a single profile is effectively national.

When DNA Is Taken: Conviction, Arrest, Detention, and “Voluntary” Collection

All 50 states, the District of Columbia, and the federal government provide for DNA collection from individuals convicted of felony offenses or other qualifying offenses. This has been the baseline since the late 1990s.

Thirty-one states, the District of Columbia, and the federal government now authorize DNA collection from individuals who have been arrested for, but not convicted of, designated offenses. The qualifying offenses vary significantly. Some states, including Alabama, California, and Kansas, collect DNA from anyone arrested for any felony. Others limit collection to specific categories of violent crime or certain enumerated offenses. A growing number of states have extended collection to certain misdemeanors as well, most commonly sex offenses and domestic violence. According to a November 2024 report prepared in partnership with Idaho State Police Forensic Services, 94.3 percent of the states that require DNA collection for misdemeanors include at least one category of sexual offense.

Federal collection authority extends further still. Under 28 C.F.R. Part 28, any federal agency that arrests or detains individuals must collect DNA samples from those who are arrested, facing charges, or convicted, as well as from non-U.S. persons detained under federal authority. This means that ICE, Customs and Border Protection (“CBP”), the U.S. Marshals Service, and other agencies are all required to collect DNA as part of their processing procedures. Between 2010 and 2020, the Department of Homeland Security operated under an exemption from this requirement, but that exemption was lifted. Consequently, DNA collection from immigration detainees resumed in 2020.

Beyond mandatory collection, police also obtain DNA through what are characterized as voluntary encounters. Officers may ask individuals to provide a cheek swab during a stop, an interrogation, or a custodial interaction. They may collect DNA from discarded items such as cigarette butts, drinking cups, or utensils under the legal theory that a person has abandoned any privacy interest in materials they have thrown away. These methods operate largely outside the statutory frameworks that at least nominally govern arrest-stage collection, and they will be examined in detail later in this article.

What Happens After Collection: Analysis, Upload, Retention, Searching, Expungement

Once a cheek swab is taken, it is sent to an accredited forensic laboratory for processing. The laboratory extracts DNA from the collected cells, measures the quantity available, and amplifies specific regions using a technique called polymerase chain reaction, or PCR. The amplified DNA is then separated and analyzed to produce the STR profile. According to the NIJ, the entire process from extraction to profile generation can take weeks or months in traditional laboratory settings, though newer Rapid DNA instruments can produce a CODIS-compatible profile from a buccal swab in approximately 90 minutes.

After the profile is generated, it is uploaded to the relevant DNA index system. A state laboratory uploads to its SDIS, from which qualifying profiles flow upward to NDIS. Once in the national system, the profile is automatically searched against the forensic index, which contains profiles from unsolved crime-scene evidence. It is also searched against other offender and arrestee profiles. These searches recur. As new forensic profiles are added from new crime scenes, they are compared against the existing database, including the profiles of people arrested years earlier. The practical consequence is that a person’s DNA profile does not sit passively in storage. It is an active participant in an ongoing, automated search process.

Retention rules vary widely. In some states, profiles and samples are retained until and unless the individual successfully completes an expungement process. In others, the state is supposed to initiate destruction automatically upon acquittal or dismissal. The FBI’s own expungement policy, governing profiles in the national index, requires the individual to submit a written request to the FBI Laboratory in Quantico, Virginia, accompanied by a certified copy of a final court order establishing that the charge was dismissed, resulted in an acquittal, or that no charge was filed within the applicable time period. This must be provided for each charge that could have served as the basis for inclusion.

Among the 31 states and the District of Columbia with active arrest-stage collection laws, the majority use a petition-based expungement model that places the burden of initiating removal on the individual. A smaller number use automatic or largely state-initiated models, but the overall pattern still favors burdening the individual. Only Michigan and South Carolina clearly place the burden on the state. Oklahoma uses an automatic model contingent on available funding. Texas uses a court-triggered model, and Utah uses a mixed model. Even where removal is not purely petition-based, the process still depends on reliable communication among courts, prosecutors, and laboratories. When those links break down, profiles remain in the system by default. The structural asymmetry is clear. Collection is swift and mandatory, while removal is slow, uncertain, and often dependent on the efforts of the very person whose rights are at stake.

Why DNA Is Not Just Another Fingerprint

The legal and political framework for pre-conviction DNA collection rests heavily on an analogy: that taking a cheek swab is essentially the same as taking a fingerprint. The Supreme Court’s majority opinion in King described DNA collection as a routine booking procedure comparable to fingerprinting and photographing. Legislators who sponsor expansion bills frequently invoke the same comparison. The analogy is central to the system’s public legitimacy. It is also misleading in several fundamental respects.

A fingerprint is a surface pattern. It identifies a person, and that is essentially all it does. A DNA sample contains a person’s complete genetic code. Even the limited STR profile used in CODIS carries more discriminating power than a fingerprint, and the underlying biological sample, which is typically retained, contains vastly more. As NYU law professor Erin Murphy observed in the context of the recent ICE collection cases, DNA reveals information about ancestry, heritable health risks, and family connections in ways that fingerprints simply cannot. A fingerprint left at a crime scene implicates the person who left it. A DNA profile in a searchable database can generate investigative leads not only against the profiled individual but against that person’s parents, children, and siblings through a technique known as familial searching.

The investigative uses are also different. Fingerprints are compared one-to-one. A latent print from a crime scene is checked against a known print. DNA profiles are searched one-to-many across a massive, continuously updated database. Every new piece of crime-scene evidence is automatically run against every stored profile in the system. A person whose DNA was collected at a protest in 2009 or a traffic stop in 2024 becomes a permanent subject of that rolling search, not because of any new evidence connecting them to criminal activity, but because their profile is there.

Finally, the retention of the biological sample introduces a dimension that has no fingerprint equivalent. A stored fingerprint card cannot be reprocessed to reveal new information about the person. A stored DNA sample can be re-analyzed as technology advances. Research has already demonstrated that STR data can, under certain conditions, be used to infer additional genetic information beyond what the original analysis was designed to capture. The sample is not a static record. It is a biological resource whose informational yield can grow over time.

None of this means that DNA evidence is without value or that forensic databases serve no public purpose. But the fingerprint analogy papers over the differences that matter most: the depth of the information captured, the breadth of the search capability, the exposure of family members who were never arrested, and the permanence of the biological sample. Understanding those differences is essential to evaluating whether the constitutional and legislative framework governing this system is adequate to the power it has created.

How the State Won the Power to Collect DNA Before Conviction

The system described in the preceding section did not emerge from a single legislative act or a single court ruling. It grew incrementally, through a combination of state legislation, federal incentives, and judicial decisions that gradually moved the constitutional boundary from conviction to arrest. But one decision more than any other shifted the legal center of gravity. In 2013, the Supreme Court’s five-to-four ruling in King declared that collecting DNA from a person arrested for a serious offense was a reasonable search under the Fourth Amendment. That decision did not create arrest-stage DNA collection, which several states had already adopted. What it did was constitutionalize the practice, turning an unresolved experiment into a settled permission from the highest court in the land.

The Road to Expansion Before King

Louisiana enacted the first state arrestee DNA collection law in 1997. Four more states followed before Congress passed the DNA Fingerprint Act of 2005, which required DNA collection from anyone arrested for a federal crime beginning in January 2009. By the time the Supreme Court took up the question, 28 states and the federal government had adopted some form of pre-conviction collection. The legal challenges that accompanied this expansion produced conflicting results. Virginia’s Supreme Court upheld its law in 2007. A California appellate court struck down that state’s law. The Maryland Court of Appeals, in the very case that would become King, found arrest-stage DNA collection unconstitutional under the Fourth Amendment. Federal and state courts had reached no consensus.

Running alongside these judicial disputes was a political and institutional momentum that made expansion increasingly difficult to reverse. The Katie Sepich Enhanced DNA Collection Act, named for a New Mexico college student whose murder was eventually linked to an offender through DNA, was first introduced in 2010 and signed into law in January 2013. It created federal incentives for states to adopt minimum collection standards for people arrested for serious violent crimes. State legislatures framed expansion as common-sense crime-fighting, and the number of states with arrest-stage laws continued to climb even as courts debated their constitutionality.

King and the Fingerprint Analogy

The facts of King gave the government its strongest possible case. Alonzo King was arrested in 2009 in Wicomico County, Maryland, on charges of first- and second-degree assault for menacing a group of people with a shotgun. During routine booking, officers collected a cheek swab under Maryland’s DNA Collection Act, which authorized sampling from individuals charged with crimes of violence or burglary. King’s DNA profile was uploaded to the state database and, weeks later, matched to DNA recovered from an unsolved 2003 rape. He was indicted, tried, and convicted. The Maryland Court of Appeals reversed, holding that the warrantless, suspicionless collection of King’s DNA violated the Fourth Amendment.

The Supreme Court reversed in turn. Justice Kennedy’s majority opinion, joined by Chief Justice Roberts and Justices Thomas, Breyer, and Alito, acknowledged that a buccal swab is a search under the Fourth Amendment. But the Court applied a totality-of-the-circumstances reasonableness test rather than requiring a warrant or individualized suspicion. Kennedy framed the collection as serving five government interests: (1) properly identifying the arrestee and determining his criminal history, (2) ensuring that custody does not create safety risks for facility staff and other detainees, (3) ensuring the arrestee would appear for trial, (4) assessing the danger he might pose to the public for purposes of bail, and (5) the possibility that identifying an arrestee as the perpetrator of an unsolved crime could help exonerate someone wrongfully imprisoned. Central to the opinion was the analogy to fingerprinting and photographing, procedures the Court described as settled and routine elements of the booking process. DNA collection, Kennedy wrote, was similarly a legitimate identification procedure that was reasonable when officers arrested someone on probable cause for a serious offense.

The opinion leaned heavily on the claim that CODIS profiles contain only non-coding STR markers that do not reveal genetic traits or medical information and that the physical intrusion of a cheek swab is negligible. It also pointed to Maryland’s statutory safeguards, including limits on when a charged person’s sample could be tested or entered into the database, the requirement that the sample be destroyed immediately if all qualifying charges were found unsupported by probable cause, and the restriction of database searches to law-enforcement identification purposes.

The Dissent’s Warning and the Missing Limiting Principle

Justice Scalia, joined by Justices Ginsburg, Sotomayor, and Kagan, filed a dissent that he read from the bench, a step traditionally reserved for the deepest disagreements. The dissent’s central argument was categorical. The Fourth Amendment forbids searching a person for evidence of a crime when there is no basis to believe the person is guilty of that crime or possesses incriminating evidence. That prohibition, Scalia wrote, applies without exception. Every prior case in which the Supreme Court had permitted a suspicionless search had required a justifying purpose apart from ordinary criminal investigation.

Scalia then dismantled the majority’s identification rationale on its own terms. He pointed out that King’s DNA sample was not analyzed until months after his arrest, that it was compared against the unsolved-crimes index rather than a database used to verify identity, and that Maryland law actually prohibited processing the sample before arraignment. The timing alone, he argued, proved that the true function of the collection was to investigate whether the arrestee had committed other offenses. The majority’s characterization of this as identification, Scalia concluded, strained credulity. The Act itself prescribed as its purpose what the Court’s own suspicionless-search precedents forbid: official investigation into a crime.

The dissent also identified the decision’s missing limiting principle. The majority confined its holding to arrests supported by probable cause for a “serious offense,” but the opinion offered no workable definition of that term and no constitutional mechanism for preventing further expansion. Scalia predicted that the logic of King would be extended to any arrest, for any offense. That prediction has proven accurate. Numerous states now authorize DNA collection for all felony arrests, and some have expanded collection to misdemeanor offenses. The majority’s “serious offense” language has functioned as a rhetorical boundary rather than a constitutional one.

What Courts Accepted and What They Sidelined

To reach its result, the majority in King had to minimize several dimensions of what DNA collection actually involves. The opinion treated the informational content of a DNA profile as limited and static, without seriously grappling with the retention of the biological sample or its potential for future re-analysis. It described the cheek swab as a minor physical intrusion comparable to fingerprinting, without addressing the qualitative difference in the information captured. And it accepted the government’s assurance that statutory safeguards would constrain future use, without examining how reliably those safeguards operated in practice.

The opinion also left unresolved how the framework would apply to collection regimes less protective than Maryland’s. Its law was among the narrower statutes, applying to crimes of violence or burglary, including attempts, and limiting testing or database entry until a probable-cause determination, a warrant-based arrest, the filing of an information, the return of an indictment, or the arrestee’s limited request or consent. California’s law, by contrast, applied to all felony arrests and lacked comparable protections. When the Ninth Circuit considered California’s regime in Haskell v. Harris, 745 F.3d 1269 (9th Cir. 2014) (en banc), following King, the en banc court upheld the law in a fractured decision, with multiple judges writing separately to express doubt about whether King’s reasoning could sustain a statute so much broader than the one the Supreme Court had reviewed.

How the Ruling Normalized Broader Collection Regimes

Whatever doctrinal limits King might have contained, its practical effect was to end the constitutional debate over arrest-stage DNA collection. Before 2013, legislatures considering expansion faced genuine legal uncertainty. Several state courts had struck down their laws, and the outcome at the Supreme Court was unknown. After King, the constitutional question was settled at the federal level, though not in every state. The number of jurisdictions authorizing arrest-stage collection grew substantially but not without setbacks. Minnesota and Vermont saw their arrestee provisions struck down under their state constitutions. As of this writing, 31 states and the District of Columbia maintain active arrest-stage authority. The federal government expanded its own collection to immigration detainees. Legislatures moved from limited offense categories to all-felony mandates with little fear of federal judicial reversal.

The ruling functioned, in practice, as a permission structure. It told legislatures that pre-conviction DNA collection would survive Fourth Amendment review, and it told courts that the reasonableness framework, rather than a warrant or individualized-suspicion requirement, would govern challenges to these laws. In the years since, no major appellate court has struck down an arrestee DNA collection statute on Fourth Amendment grounds. The question that remains open is not whether the government can collect DNA before conviction but whether any constitutional limit exists on how broadly, how permanently, and for what purposes it may use what it has collected.

The Official Justification

The arguments for expanding pre-conviction DNA collection are not trivial. They have persuaded the legislatures of more than 30 states, a majority of the Supreme Court, and both major political parties. Before examining where those arguments fall short, this section presents them at their strongest, because an article that caricatures the opposing case weakens its own credibility.

Solving Cold Cases

The most emotionally powerful argument for expansion is that a broader DNA database connects more suspects to unsolved violent crimes. Advocates regularly cite cases in which an arrest-stage DNA match solved a rape or murder that had gone cold for years. The facts of King itself provided a vivid example. Alonzo King’s cheek swab at booking linked him to a 2003 rape that investigators had been unable to solve for six years. In Ohio, a DNA match on the second day of that state’s arrest-stage law connected a felony abduction suspect to a decade-old rape. A 2016 Indiana homicide, in which an 82-year-old man was shot while checking his mailbox, was cracked only because the suspect’s DNA had been collected during an earlier Ohio robbery arrest that never resulted in a conviction. These cases are real, documented, and consequential. They carry substantial weight in legislative hearings because they involve identifiable victims whose suffering could have been shortened by earlier identification of the offender.

Identifying Repeat Offenders Earlier

A related argument holds that earlier DNA collection intercepts serial offenders before they reoffend. If a person is arrested for a property crime and their DNA matches evidence from a prior sexual assault, the database has functioned as an early-warning system. Proponents such as Jennifer Doleac of Arnold Ventures have argued that individuals whose profiles are added to the database are significantly less likely to reoffend, suggesting a deterrent effect. The logic is straightforward. If offenders know their genetic profile is in the system, they face a higher probability of being identified if they commit future crimes. Legislators in Missouri and Iowa cited this rationale prominently during their 2026 expansion debates, framing broader collection as a tool for preventing violence rather than merely documenting it after the fact.

Exonerating the Innocent

Supporters also emphasize that DNA databases can work in reverse. When a crime-scene profile matches an individual already in the system, it can exclude other suspects who might otherwise face prosecution. The National Registry of Exonerations has documented hundreds of cases in which DNA evidence contributed to overturning wrongful convictions. Expansion advocates argue that a larger database increases the chances that the actual perpetrator is identified, reducing the risk that an innocent person is convicted based on weaker evidence. Ashley Spence, founder of the DNA Justice Project and a survivor of sexual assault, told Missouri legislators in 2026 that broader collection protects those who are wrongfully accused by ensuring that the right person is identified through science rather than circumstantial evidence. This exoneration argument matters because it allows supporters to frame expansion as serving both public safety and individual justice.

Helping Victims and Improving Clearance Rates

Victim-centered rhetoric occupies a central place in the legislative case for expansion. Supporters argue that victims of violent crime deserve every available investigative tool and that restricting DNA collection to post-conviction leaves a critical gap during which offenders remain unidentified and free to harm others. Clearance rates for violent crime have declined in many jurisdictions. In Missouri, only 38 percent of violent crimes reported to law enforcement resulted in clearance in 2023, according to data cited during the state’s legislative debate. Proponents of expansion contend that larger DNA databases directly address this gap by generating investigative leads that would otherwise never materialize. When supporters frame the issue in these terms, opposing expansion can appear tantamount to indifference toward victims of violent crime, a framing that has proven politically effective in state after state.

The Political Appeal of “Just One More Tool”

Perhaps the most important feature of the expansion argument is how modest it sounds. DNA collection is presented as incremental, technical, and common-sense. Legislators describe it as giving law enforcement “one more tool” rather than as a structural expansion of state power over the unconvicted. The physical act of collection, a cheek swab lasting seconds, reinforces this framing. The analogy to fingerprinting, affirmed by the Supreme Court, provides constitutional cover. And the existence of expungement provisions, however they function in practice, allows supporters to assure skeptics that the innocent can have their records removed.

This incremental framing has proven effective across partisan lines. Conservative legislators support expansion as a law-and-order measure. Progressive legislators support it as a tool for protecting victims and improving forensic accuracy. The bipartisan appeal helps explain why the number of states with arrest-stage DNA laws has grown steadily for two decades with relatively little organized political opposition. The question the next section examines is whether the strength of the justification matches the breadth of the power it has been used to authorize.

Why the Official Story Is Incomplete

The arguments presented in the preceding section are genuine. DNA databases have solved cold cases. They have identified serial offenders. They have, in documented instances, contributed to the exoneration of wrongly accused individuals. None of that is in dispute. What is in dispute is whether those outcomes justify the particular system that has been built to achieve them: one that collects, retains, and continuously searches the genetic profiles of millions of people who have never been convicted of a crime.

Clearance Rhetoric vs. Real-World Evidence

Legislative debates over DNA expansion are dominated by specific, emotionally compelling cases. A rape is solved after a decade, a serial predator is identified through a booking swab, a murder is cracked because a suspect’s profile was already in the system. These stories are powerful precisely because they are concrete. But the question a governing principle must answer is not whether a broader database can produce individual successes. It is whether dragnet-style collection from all arrestees produces systemwide benefits proportionate to the scope of the intrusion.

The evidence on that broader question is thinner than the rhetoric suggests. The FBI’s own statistics show that CODIS has aided more than 758,000 investigations since the system’s inception, but that figure encompasses all database categories, including convicted-offender profiles, which constitute the vast majority of stored records. Proponents rarely isolate the marginal contribution of arrestee profiles specifically. When the ACLU challenged California’s arrest-stage collection law, it argued, citing the government’s own data, that DNA collected from arrestees did not help solve crimes at a rate measurably higher than DNA collected only from those who were actually convicted. The question is not whether databases work but whether the leap from convicted-offender databases to arrestee databases delivers public-safety gains sufficient to justify the constitutional cost.

The Gap Between Anecdote and Governing Principle

Exceptional cases make poor foundations for rules of general application. The cold-case hits that dominate legislative testimony involve specific, identifiable offenders whose prior arrest happened to precede their identification for a separate, serious crime. But arrest-stage DNA laws do not apply only to those people. They apply to everyone swept into the system on a qualifying charge, regardless of whether that charge leads to prosecution, conviction, or anything at all. A policy justified by its capacity to catch the occasional serial rapist also captures the person arrested at a protest who is never charged, the young man picked up on a drug possession allegation that is later dismissed, and the domestic-violence suspect arrested after a disputed incident and released the same day.

The scale of non-conviction outcomes matters here. Bureau of Justice Statistics data indicate that approximately one-third of state felony cases do not result in conviction. In California, roughly 22 percent of felony arrests are dropped before charges are even filed with a court. Justice Scalia’s dissent in King noted that nearly one-third of Americans will be arrested for some offense by age 23. The database is not populated exclusively, or even primarily, by the offenders whose stories appear in legislative testimony. It is populated by the full breadth of people who pass through the arrest stage of a criminal legal system defined by high-volume processing, discretionary charging, and frequent non-conviction outcomes.

Arrest Is Not Guilt

The presumption of innocence is not a slogan. It is a substantive constitutional principle that allocates the burden of proof to the government and forbids the state from imposing the consequences of guilt on people who have not been found guilty. Pre-conviction DNA collection inverts that principle at the biological level. A person who is arrested, processed, and released without conviction has been subjected to a permanent genetic capture on the strength of an accusation alone. Their profile enters a searchable national database and remains there unless they successfully navigate an expungement process that, as earlier sections have shown, is slow, burdensome, and frequently ineffective.

The state’s position, ratified by King, is that this does not offend the presumption of innocence because collection is merely an administrative booking step and because expungement is available. But the administrative label does not change the substance of what occurs. And the availability of expungement in theory does not alter the reality that, in practice, large numbers of unconvicted individuals remain in the database indefinitely. The system treats accusation as sufficient grounds for lasting biological mass surveillance. That is a significant departure from the principle that citizens are free from the investigative apparatus of the state until the state proves its case.

Permanent Collection in a System Built on Nonfinal Accusations

Pre-conviction DNA collection would raise constitutional concerns even in an idealized criminal legal system. It raises far graver concerns in the one that actually exists. American criminal justice is characterized by broad police discretion over whom to arrest, heavy reliance on plea bargaining to resolve cases, and enormous variation in how charges are filed, reduced, or dismissed. Arrest is not a careful judicial determination of probable involvement in serious criminal activity. It is an initial, often low-threshold action by a police officer in the field, subject to error, bias, and strategic calculation.

In that system, tying permanent DNA capture to the arrest stage means tying it to the point where state power is at its most discretionary and least accountable. Communities subject to heavier policing generate more arrests, more DNA collections, and more permanent database entries, regardless of whether those arrests reflect higher rates of criminal conduct or simply higher rates of police contact. The result is a database whose composition reflects policing patterns as much as it reflects crime patterns. This dynamic will be examined more fully in a later section, but it is essential to flag here. The constitutional problem with pre-conviction DNA collection is inseparable from the institutional reality of the system in which it operates.

How Temporary Exceptions Become Ordinary Police Power

The expansion of DNA collection follows a pattern visible across many areas of law enforcement authority. A power is introduced as a narrow exception, justified by urgent circumstances. It is framed as limited, targeted, and subject to safeguards. Over time, the exception widens. The safeguards erode or prove unworkable. What was presented as extraordinary becomes routine.

Arrest-stage DNA collection has traced exactly this arc. It began with a handful of states authorizing collection for a limited set of violent felonies. It expanded to all felonies, then to certain misdemeanors. Federal collection, initially limited to convicted offenders, was extended to arrestees in 2009, then to immigration detainees in 2020, and, according to recent reporting, to protesters arrested during political demonstrations. At each stage, the expansion was presented as modest and incremental. At no stage was a clear stopping point articulated. The question the next sections address is what this trajectory means in practice: for the reliability of DNA evidence, for the communities most affected, and for the constitutional values the system was supposed to protect.

DNA Is Powerful Evidence but Not Infallible Evidence

The public prestige of DNA evidence is enormous. Jurors, judges, and legislators tend to treat it as uniquely objective and self-interpreting, a form of proof that speaks for itself. In many contexts, that reputation is earned. Single-source DNA analysis, properly conducted, is among the most reliable forensic methods available. But a great deal of the DNA evidence that enters the criminal legal system does not come in that clean, single-source form. It arrives as trace material, degraded samples, and complex mixtures that require layers of interpretation before they mean anything at all. Understanding those layers is essential, because the expansion of DNA collection magnifies not only the power of the evidence but also the consequences of its misuse.

Transfer DNA, Secondary Transfer, and Trace Evidence Problems

Since 1997, when Australian forensic scientist Roland van Oorschot published research showing that DNA could be detected from the mere touch of a hand, forensic science has grappled with a problem it did not initially anticipate. DNA can appear on objects and in places without any direct contact by the person whose genetic material is found. An average person sheds tens of millions of skin cells daily. A handshake can deposit enough cellular material to generate a detectable profile. And that deposited material can travel further. A person touches a doorknob, a second person touches the same doorknob, and the first person’s DNA ends up on the second person’s hand, or clothing, or in a location they have never visited.

The case of Lukis Anderson illustrates the danger with devastating clarity. In 2012, Anderson, a 26-year-old homeless man in San Jose, California, was charged with the murder of a Silicon Valley investor named Raveesh Kumra. Anderson’s DNA had been found under Kumra’s fingernails. The physical evidence appeared damning. But Anderson had been hospitalized in a near-comatose state with a blood alcohol level five times the legal limit on the night of the killing. Medical records confirmed he had been in the hospital the entire time. The explanation, pieced together months later by his public defender, was that paramedics who had treated Anderson earlier that evening subsequently responded to the Kumra residence. A pulse oximeter or other equipment carried Anderson’s DNA from one patient to the other. He spent nearly six months in jail on a murder charge before the case collapsed. Anderson was charged with a capital crime on the strength of DNA evidence that proved nothing about his involvement.

Contamination, Mixtures, and Low-Template Ambiguity

Even when transfer is not at issue, DNA evidence can be far more ambiguous than it appears. Contamination during collection, transport, or laboratory processing can introduce foreign DNA into a sample. Crime-scene evidence frequently contains mixtures of genetic material from multiple individuals, and separating those contributions is a task that involves significant analytical judgment. When the amount of DNA recovered is extremely small, the resulting profile may be partial or difficult to interpret with confidence.

In 2016, the President’s Council of Advisors on Science and Technology (“PCAST”) published a landmark report evaluating the scientific validity of several forensic comparison methods. PCAST found that single-source DNA analysis and simple two-person mixtures met its standards for foundational scientific validity. But it concluded that subjective analysis of complex DNA mixtures, those involving three or more contributors, had not been established as a reliable methodology. The report noted that subjective choices made by examiners during the interpretation of complex mixtures could dramatically affect the results, introducing significant risk of analytical error and confirmation bias. Complex mixtures constitute a meaningful share of the DNA evidence generated in criminal investigations, and they are precisely the kind of evidence most likely to arise from the trace-DNA scenarios that expanded collection makes more common.

The Danger of Overstating What a Match Means

When a forensic laboratory reports that a suspect is “included as a possible contributor” to a DNA sample, it is making a statistical statement about whether the suspect’s genetic profile is consistent with the evidence. It is not making a statement about whether the suspect was at the crime scene, touched the weapon, or committed the offense. That distinction collapses routinely in courtroom presentation. Prosecutors, expert witnesses, and even judges regularly describe DNA matches in language that implies certainty about conduct rather than compatibility with a profile. A match becomes “proof” that the defendant was present. Presence becomes proof of participation.

The cultural authority of DNA compounds this problem. Research has found that jurors assign extraordinary weight to DNA evidence, often treating it as dispositive regardless of what the rest of the evidence shows. When the underlying sample involves trace material, a complex mixture, or an ambiguous statistical result, that deference can be dangerous. A profile that is merely consistent with a suspect’s inclusion, out of a population of thousands or millions of possible contributors, is not the same as a definitive identification. But it is frequently presented, and received, as if it were.

Probabilistic Genotyping and Black-Box Interpretation

Recognizing the problems with subjective mixture interpretation, the forensic community has increasingly turned to probabilistic genotyping software, programs that use statistical algorithms to assess the likelihood that a given individual contributed to a mixed DNA sample. PCAST described these tools as a significant improvement over purely subjective methods. But it also warned that the software required careful scrutiny. The council found that validity had been established only under limited conditions, specifically for three-person mixtures in which the minor contributor constituted at least 20 percent of the total DNA. For more complex or degraded samples, substantially more empirical testing was needed.

Defense attorneys and independent scientists have raised additional concerns. The leading probabilistic genotyping programs, STRmix and TrueAllele, are proprietary. Their source code is not routinely disclosed to defendants. In at least one case documented in the PCAST addendum, the two programs reached different conclusions about whether a defendant’s DNA was present in a mixture. When the tools that interpret the evidence are opaque to the people against whom that evidence is used, due process is diminished regardless of how sophisticated the underlying mathematics may be.

Why Juries and Courts Often Over-Credit DNA Evidence

DNA’s reputation was built on its most dramatic successes: cold-case identifications, wrongful-conviction exonerations, and single-source matches with vanishing probability of error. Those successes are real. But they have created a halo effect that extends to all DNA evidence, including the trace samples, partial profiles, and complex mixtures where the science is far less certain. Jurors who hear that a defendant’s DNA was found on a piece of evidence may not appreciate that the finding is consistent with incidental transfer, laboratory contamination, or a statistical model that includes thousands of other possible contributors.

The expansion of pre-conviction DNA collection feeds directly into this dynamic. A larger database generates more matches. More matches produce more cases in which DNA evidence is central to prosecution. And in each of those cases, the cultural weight of DNA can magnify the consequences of every interpretive error, every overstated probability, and every failure to disclose the limitations of the analysis. The problem is that the system treats it as though it is without flaw, and that assumption becomes more consequential with every profile added to the database.

From Identification to Investigation to Surveillance

The legal framework that authorized pre-conviction DNA collection was built on the claim that the process serves an identification function. The Supreme Court in King described it as helping officers determine who they have in custody. Legislators have consistently framed it as a booking procedure, a formality that happens alongside fingerprinting and photographing. But the operational reality of the system that has grown around that framing bears little resemblance to the modest identification tool it was sold as. Once a profile enters CODIS, it does far more than confirm a name. It becomes a permanent investigative asset, searchable across jurisdictions, usable against relatives, and retained long after the legal basis for the original collection has disappeared.

DNA as More Than Identification

Identification, in the ordinary sense of the word, means confirming who someone is. Fingerprinting does this. Photographing does this. A DNA profile in CODIS does this too, but it also does something qualitatively different. It connects the profiled individual to evidence from unsolved crimes they may have had no known connection to at the time of collection. That is investigation, not identification. When Alonzo King’s cheek swab was compared against the unsolved-crimes index and matched to a 2003 rape, the system was not verifying his name. It was generating a new criminal lead. The entire value proposition of arrest-stage DNA collection depends on this investigative function. Cold cases are not solved by confirming a suspect’s identity at booking. They are solved by running a profile against crime-scene evidence from unrelated cases. The system is designed for investigation, despite the fact it is described as identification.

Familial Searching and the Widening Circle of Suspicion

One person’s inclusion in a DNA database does not affect only that person. Because close biological relatives share significant portions of their genetic material, a partial match between a crime-scene sample and a stored profile can generate investigative leads pointing to the profiled individual’s parents, children, and siblings. This technique, known as familial searching, is practiced in at least 10 states, including California, Colorado, Texas, and Virginia. The FBI does not conduct familial searches at the national level, but it does not prohibit states from conducting them within their own databases. Maryland and the District of Columbia have passed laws specifically banning the practice.

Familial searching means that the genetic surveillance created by arrest-stage DNA collection extends beyond the individual whose cheek was swabbed. A person who was arrested at a protest and never charged may find that their presence in the database has drawn their brother or their son into a criminal investigation. The Federal Judicial Center has noted that given the disproportionate representation of racial minorities in law-enforcement DNA databases, certain individuals face a higher risk of being investigated simply because genetic relatives are already in the system. The circle of suspicion radiates outward from every stored profile, and as the database grows, so does the number of people who can be drawn into that circle without ever having been arrested themselves.

Retention After Non-Conviction

Earlier sections described the structural asymmetry between collection and removal. That asymmetry has a direct surveillance consequence. When a person is arrested, swabbed, and released without conviction, their profile does not automatically vanish from the system in most jurisdictions. In the majority of arrest-stage jurisdictions, the individual must affirmatively seek removal by submitting documentation to the relevant agency. Federal expungement requires a written request to the FBI Laboratory accompanied by certified court orders for each qualifying charge. Many people do not know these procedures exist. Those who do may lack the resources, legal knowledge, or access to courts necessary to complete them.

The practical result is that profiles from non-conviction cases accumulate in the database. Each one remains searchable against every new piece of crime-scene evidence uploaded to the system. A person arrested in 2015 whose charges were dropped in 2016 may still be subject to automated database comparisons a decade later. The collection was justified as a temporary incident of booking. The retention transforms it into a lasting investigative resource.

Cross-Agency Sharing and Database Growth

CODIS is a shared system. Its three-tiered architecture was designed to enable cross-jurisdictional searching. Profiles flow from local to state to national levels, and a cheek swab taken by a county sheriff’s office can become accessible to federal investigators and vice versa. Federal law permits disclosure of stored DNA samples and analyses to criminal justice agencies for law-enforcement identification purposes, in judicial proceedings if otherwise admissible, to criminal defendants for defense purposes in their own cases, and – if personally identifiable information is removed – for population statistics, identification research and protocol development, or quality control. The 2020 expansion of federal DNA collection to immigration detainees added an entirely new population to the system. Reports in 2026 revealed that DNA was being collected from protesters arrested by ICE officers during political demonstrations.

Each expansion adds profiles. Each new set of profiles increases the pool against which future crime-scene evidence is compared. And the bureaucratic momentum of a large, integrated database creates its own justification for continued growth. Agencies that have invested in collection infrastructure, laboratory capacity, and interoperability have institutional incentives to maintain and expand the system regardless of whether the original policy rationale still holds.

When Booking Becomes a Long-Term Investigative Asset

This is the central mismatch the article has been building toward. At the front end, DNA collection looks like a booking procedure that entails a brief swab, a few seconds of physical contact, and a cotton-tipped applicator sealed in an envelope. At the back end, that swab generates a digital profile that enters a national database, is automatically and continuously searched against evidence from unsolved crimes across the country, can expose the profiled individual’s biological relatives to investigation, and persists indefinitely unless the individual successfully navigates a removal process designed to be difficult.

What is described as a moment becomes a structure. What is framed as identification functions as investigation. What is presented as temporary operates as permanent. The gap between the rhetoric and the reality is not a minor discrepancy. It is the central constitutional problem with the system as it exists today. And the people who bear its consequences most heavily are not the serial offenders whose cases dominate legislative hearings. They are the millions of unconvicted individuals whose genetic profiles sit in a searchable law-enforcement database because they were once arrested, for any reason, and the system collected what it could.

How Police Obtain DNA Without Meaningful Consent

The preceding sections have traced the legal and institutional architecture of arrest-stage DNA collection. But a significant share of DNA in law-enforcement hands was never collected through that statutory framework at all. Across the country, police obtain genetic material through encounters that fall outside the booking process, using methods that lack the procedural protections even the current system nominally provides. Officers ask people on the street to submit to cheek swabs during stops that do not result in arrest. They collect saliva by offering suspects a drink during custodial questioning. They swab minors without notifying a parent. And they harvest DNA from discarded objects under the legal theory that a person who throws away a coffee cup has abandoned any privacy interest in the genetic information it carries.

These practices operate in a space between coercion and consent where the distinction between the two largely disintegrates. The person asked to provide a sample during a police encounter has not been told they must comply, but neither have they been told they can refuse without consequence. The interaction carries the inherent authority of a uniformed officer, the implied threat that refusal may invite further scrutiny, and the situational pressure of a stop, a detention, or an interrogation. What the law characterizes as voluntary often bears little resemblance to a free and informed choice.

“Voluntary” Collection in the Field

In 2016, ProPublica published an investigation documenting how police departments in Florida, Pennsylvania, and elsewhere had begun collecting DNA from people who were neither arrested nor suspected of any particular crime. In Melbourne, Florida, five teenage boys sitting in a parked car in a gated community were approached by an officer who told them there had been recent car break-ins in the area. After questioning them, the officer asked which one wanted to provide a DNA sample. A 15-year-old named Adam agreed. He rubbed a swab along the inside of his cheek, signed a consent form, and gave his thumbprint. No parent was present. No parent was contacted. Under the Melbourne Police Department’s policy, the boy’s agreement was sufficient.

ProPublica’s investigation found that this was not an isolated incident. The Melbourne Police Department collected 307 cheek swabs in 2015. Thirty-eight of those were taken during field interviews unconnected to any particular crime. Two nearby departments – Palm Bay and West Melbourne – acknowledged that they had asked minors to provide DNA without parental permission. In Bensalem Township, Pennsylvania, people stopped for DUI or for acting “suspiciously” were routinely asked for cheek swabs. The resulting profiles were stored in a local database shared among 39 police departments across Bucks County, funded with drug forfeiture money and processed by a private forensic laboratory. None of this was governed by the statutory frameworks that regulate state or federal DNA databases.

The Washington University Law Review documented a broader pattern in a 2023 article examining what the author termed “DNA on Demand.” According to the article, the New York Police Department maintained a practice of secretly collecting DNA by offering suspects a drink during custodial interrogations, then using the discarded container to extract a genetic profile. The resulting samples were added to a local index maintained by the Office of the Chief Medical Examiner. That index held approximately 64,000 profiles by 2017, according to reporting by The Trace and WNYC, and was subject to no published rules governing whose DNA could be entered, how long it could be retained, or under what circumstances it could be searched.

Discarded DNA and the Abandonment Doctrine

When police cannot persuade someone to submit a cheek swab, they can often obtain the same genetic material without asking at all. Under the legal doctrine of abandonment, a person who discards an object in a public place is often held to have relinquished any reasonable expectation of privacy in it. Courts have applied this principle to allow police to collect DNA from cigarette butts left on a sidewalk, coffee cups thrown into a trash can, and utensils discarded after a meal. The Fourth Amendment, as currently interpreted, does not require a warrant for these collections because the individual is deemed to have voluntarily parted with the material.

The abandonment doctrine was developed for physical objects. Applying it to genetic material produces consequences the doctrine was not designed to address. A person who throws away a cup has no reason to believe they are providing the government with a sample that can identify them, reveal their ancestry and heritable health risks, and expose their family members to investigation through familial searching. The informational gap between discarding a physical object and surrendering a biological specimen is enormous, but the law treats the two as equivalent.

The Special Vulnerability
of Minors and Detainees

The coercive potential of informal DNA collection is highest when the person being asked occupies a position of particular vulnerability. Minors cannot meaningfully evaluate the long-term consequences of providing a DNA sample to police, but, as ProPublica documented, departments in Florida collected swabs from teenagers without so much as contacting a parent. The 15-year-old in Melbourne needed parental consent for a school field trip. He did not need it to give his genetic information to a police officer.

Immigration detainees face a comparable imbalance. Under federal regulations adopted in 2020, CBP and ICE are required to collect DNA from individuals held under federal authority, including asylum seekers and unaccompanied minors. A 2025 essay published by the Hastings Center reported that the government had been collecting DNA from migrant children, some as young as four years old, and storing it in the FBI’s national criminal database. These children have not been accused of any crime. Many cannot speak English. The Hastings Center described the practice as a predictable extension of a decades-long pattern of expanding genetic surveillance under the justification of security.

What unites these practices is the absence of the procedural protections that even the flawed statutory system provides. Arrest-stage collection, for all its constitutional problems, at least occurs within a defined legal framework, i.e., triggered by a specific event, governed by statute, and subject in theory to expungement. DNA obtained through street encounters, custodial manipulation, discarded-item collection, and the processing of vulnerable populations operates outside that framework entirely. There is no triggering arrest, no statutory authorization in most cases, no expungement mechanism, and no meaningful opportunity to refuse. The result is a shadow system of genetic collection that supplements the formal one, expanding the reach of law-enforcement DNA databases well beyond what any legislature has voted to authorize.

Who Bears the Burden
of Expansion

The expansion of pre-conviction DNA collection is often presented in race-neutral terms. Legislators describe it as applying equally to anyone arrested for a qualifying offense. Courts evaluate it under a Fourth Amendment framework that, by design, applies the same reasonableness standard regardless of who is searched. But the system these laws feed into is not race-neutral, and the burdens of genetic surveillance do not fall equally. Because DNA collection is tied to arrest, the demographics of the database reflect the demographics of policing. Communities subjected to heavier enforcement generate more arrests, more swabs, and more permanent database entries. The result is a system in which the genetic information of Black and Latino individuals is collected, stored, and searched at rates far out of proportion to their share of the population.

The Database Mirrors
the Disparities

In a 2020 study published in the California Law Review, NYU law professor Erin Murphy and researcher Jun Tong set out to determine the actual racial composition of forensic DNA databases. Using data obtained through freedom-of-information requests and a reverse-engineering methodology based on each state’s collection policies and demographic data, they found that DNA profiles from Black individuals were collected at two to three times the rate of White individuals nationwide. The disparity was not an accident of individual case outcomes. It was a structural consequence of tying collection to arrest in a system where arrest rates are themselves shaped by racially uneven patterns of policing, prosecution, and enforcement priority.

This finding matters because it transforms the constitutional question. If DNA databases reflected the general population, the argument that collection is a neutral administrative procedure would be easier to sustain. They do not. They reflect the same disparities that characterize the criminal legal system at every other stage, from the decision about where to deploy officers to the decision about whom to stop, whom to arrest, and whom to charge. Murphy and Tong’s data showed that the disparity was present across states with different collection policies, but it widened in states that extended collection to arrestees rather than limiting it to convicted offenders. Arrest-stage collection does not merely inherit the system’s existing inequities. It amplifies them, because arrest is the point in the process where police discretion is broadest and judicial oversight is thinnest.

Family Spillover

The racial skew of the database extends its reach beyond the individuals whose profiles it contains. As earlier sections explained, familial searching allows investigators to use a partial DNA match to identify the biological relatives of a profiled individual. When the database disproportionately contains profiles from Black and Latino men, familial searching disproportionately subjects their parents, children, and siblings to genetic investigation. The Federal Judicial Center has acknowledged that given the overrepresentation of racial minorities in law-enforcement DNA databases, certain populations face a heightened risk of investigation based on a relative’s inclusion in the system.

A person who has never been arrested, never been charged, and never had any contact with the criminal legal system can be drawn into a criminal investigation because a brother or a son was once arrested on a charge that was later dismissed. The original arrest generated a database entry. The database entry persisted because expungement did not occur. And that persistent entry created an investigative pathway to family members who had no involvement whatsoever. In communities where arrest rates are highest, this radiating effect is most concentrated. The genetic surveillance created by a single booking swab does not stop with the person swabbed.

Immigration Enforcement and the Border

Federal DNA collection from immigration detainees adds another dimension to the distributional burden. Under regulations that took effect in 2020, CBP and ICE collect DNA from individuals detained under federal authority. The overwhelming majority of people processed through the immigration enforcement system are Latino, and a growing number are Indigenous people from Central America and southern Mexico. The Duke Law Journal, in a 2025 article analyzing the constitutional gap surrounding mass biometric collection at the border, observed that the federal regulation effectively created a large-scale genetic indexing program applied almost exclusively to a single racial and ethnic population.

The Hastings Center’s 2025 bioethics essay on the collection of DNA from migrant children described the practice as disproportionately affecting Latino and Indigenous groups and warned that it risked embedding genetic profiling as a tool of systemic discrimination. These children are not criminal suspects. Many are asylum seekers or unaccompanied minors. Their inclusion in a law-enforcement database is a consequence of immigration policy, not criminal conduct, but their profiles sit alongside those of convicted offenders and felony arrestees in the same searchable system.

The Unequal Burden of Removal

Even the mechanisms designed to limit the system’s reach operate unevenly. Expungement requires navigating a bureaucratic process that presupposes access to legal knowledge, documentation, and in many states, a lawyer. Federal expungement demands a written request to the FBI Laboratory accompanied by certified court orders for each qualifying charge. These requirements fall hardest on the people least equipped to meet them – individuals without stable housing, without legal representation, and without familiarity with the administrative system they are asked to petition. In practice, the people most likely to have their DNA collected through arrest-stage laws are also the people least likely to secure its removal after non-conviction.

The asymmetry is structural. Collection is automatic, fast, and requires nothing from the individual except physical compliance. Removal is manual, slow, and requires the individual to initiate and complete a process that the system has no institutional incentive to facilitate. When that asymmetry is layered on top of the racial and socioeconomic disparities already embedded in the database, the result is a system that retains genetic profiles from disadvantaged communities at higher rates and for longer periods than from the population as a whole.

None of these distributional consequences are hypothetical. They are documented in the academic literature, acknowledged in judicial and government sources, and visible in the composition of the database itself. A system that collects genetic material based on arrest, in a country where arrest is shaped by race, geography, and economic status, will inevitably produce a database that reflects those forces. The question is whether that outcome is constitutionally acceptable.

The New Expansion Playbook in the States

The trajectory described in the preceding sections is not slowing down. In the first quarter of 2026, legislatures in Missouri, Iowa, Georgia, and Kentucky each advanced bills to expand DNA collection beyond their existing statutes. The pattern across the four states is consistent in direction even if not identical in form. That is, move collection earlier in the process and broaden the pool of people subject to compulsory sampling. Missouri sought all-felony arrest-stage collection; Iowa sought arrest-stage collection for felony and violent aggravated-misdemeanor cases; Georgia sought new booking-stage collection authority for felony arrestees; and Kentucky sought to expand collection to persons charged with a felony. What differs in 2026 is the scale of the ambition and the near-total absence of organized legislative opposition.

Missouri: From Targeted
Collection to All Felonies

Missouri’s existing law authorizes DNA collection from persons 17 or older arrested for first- or second-degree burglary or for a felony offense under Chapters 565, 566, 567, 568, or 573. In 2026, lawmakers sought to eliminate those categorical limits. Senate Bill 1458 would extend DNA collection to every person 17 or older arrested for any felony offense, and companion House Bill 2868 would do the same. As of April 1, 2026, SB 1458 had received a do-pass recommendation from the Senate Judiciary and Civil and Criminal Jurisprudence Committee, and HB 2868 had been placed on the House Informal Perfection Calendar. Neither bill had yet received a floor vote or been enacted.

Iowa: Moving from Conviction-Only to Arrest

Iowa’s 2026 proposal was more dramatic because it required a larger structural shift. Iowa had been a conviction-only state, collecting DNA only after a guilty finding. House File 2624, championed by Attorney General Brenna Bird (R) and Governor Kim Reynolds (R), proposed moving Iowa to an arrestee model for all felonies and violent aggravated misdemeanors. The bill passed the Iowa House on March 10, 2026, by a vote of 64 to 30. A fiscal analysis estimated the proposal would cost the state nearly $2 million over two years, with $1.4 million falling on the Division of Criminal Investigation in the first year. As of late March 2026, the bill was pending before the Senate Judiciary Committee.

Supporters in Iowa deployed the same rhetorical toolkit seen in Missouri: named victims, cold-case success stories, and the argument that Iowa was simply catching up with the majority of states. Opponents raised concerns about collecting DNA from people who are presumed innocent, but the 64-to-30 margin suggested those objections had limited legislative traction.

Georgia and Kentucky

Georgia’s Ashley Spence Act, Senate Bill 29, would newly authorize DNA collection from persons arrested for the commission of a felony at booking or processing. As of March 27, 2026, the bill had been recommitted to committee. Kentucky’s House Bill 414 would expand collection to persons charged with a felony. The bill passed the House and remained pending in the Senate Judiciary Committee as of April 1, 2026.

The Playbook and Its Blind Spots

Taken together, the 2026 legislative cycle reveals a coordinated expansion effort with a repeatable structure. The same advocacy organizations testify in multiple state capitals during the same session. The same victim stories anchor the testimony. The same talking points frame resistance as indifference to public safety. And the same comparison to other states’ laws creates the impression that expansion is inevitable and noncontroversial.

What the playbook does not include is any serious engagement with the constitutional and structural concerns this article has documented. Legislative hearings in Missouri, Iowa, Georgia, and Kentucky focused almost entirely on cold-case outcomes and victim advocacy. They did not address the racial composition of DNA databases, the failure of expungement mechanisms, the retention of biological samples, the expansion of familial searching, or the growing use of DNA collection against protesters and immigration detainees. The bills were debated as if the only question were whether DNA databases help solve crimes. The harder question, whether the scope of the collection is proportionate to the power it creates, was not on the agenda.

The Hidden Weakness of “Safeguards”

Every legislative debate over DNA expansion includes a moment when a sponsor addresses the civil-liberties objection. The reassurance follows a predictable script. Safeguards are in place, profiles can be removed, and samples will be destroyed if charges don’t lead to conviction. These assurances are not fabricated. Statutes do contain expungement provisions. Federal law does require laboratories to have destruction procedures. The problem is that the distance between the safeguard as written and the safeguard as experienced by the person whose DNA is in the system is enormous. And that distance is not an accident of implementation. It is built into the structure of the system itself.

The Federal Expungement Burden

The FBI’s expungement policy for profiles in the National DNA Index System illustrates the structural problem with unusual clarity. To have a federal DNA record removed, an individual must submit a written request to the FBI Laboratory at Quantico, Virginia. The request must be accompanied by a certified copy of a final court order establishing, for each qualifying charge, that the charge was dismissed, resulted in an acquittal, or that no charge was filed within the applicable time period. The policy states that it does not apply to records resulting from state convictions or arrests; individuals seeking removal of state-level records must contact the relevant state agency separately.

Each element of that process imposes a burden the person may not be able to meet. Obtaining a certified court order requires access to the court system and knowledge of how to navigate it. When multiple charges served as the basis for inclusion, a separate order is needed for each one. Many people who pass through the federal arrest-and-release cycle do not have lawyers. They may not know the policy exists. They may not know their DNA was uploaded to a national database in the first place. And if they do know, the task of assembling certified documentation months or years after the fact, from a system that keeps no record of their genetic capture on any document they receive at release, may be practically impossible.

The Petition-State Problem

The federal model is not an outlier. Among the 31 states and the District of Columbia with active arrest-stage DNA collection laws, the majority use a petition-based expungement system that places the burden of initiating removal on the individual. In California, a person with no qualifying conviction may submit a written request for expungement but only if no accusatory pleading was filed within the applicable time period or if the person was acquitted or found factually innocent. In Colorado, the Bureau of Investigation allows expungement if a felony charge was not filed within 90 days or if the case ended in dismissal or acquittal, but the individual must initiate the process.

A smaller number use automatic or largely state-initiated models, but the overall pattern still favors burdening the individual. Only Michigan and South Carolina clearly place the burden on the state. Oklahoma uses an automatic model contingent on available funding. Texas uses a court-triggered model, and Utah uses a mixed model. Even where removal is not purely petition-based, the process still depends on reliable communication among courts, prosecutors, and laboratories. When those links break down, the profile remains in the system by default.

Sample Retention

Expungement of a digital profile does not necessarily mean destruction of the biological sample. Federal law requires that expungement include destruction of the associated DNA samples, and several states impose the same requirement. But the physical sample, stored on a card or in a sealed container at the collecting laboratory, exists in a different chain of custody than the digital profile in the database. Ensuring that both are destroyed requires coordination between the database administrator and the laboratory that retains the sample. When that coordination fails, the digital profile may be removed while the biological specimen persists, preserving the possibility of future re-analysis.

The retention of biological samples is not a minor technical detail. The sample contains a person’s full genetic information, far more than the limited STR profile uploaded to CODIS. A stored sample can be re-analyzed using techniques that did not exist when it was collected. As long as the sample survives, the informational power of the original collection can grow over time, regardless of whether the digital profile has been formally expunged.

The Structural Asymmetry

The pattern across all of these mechanisms is the same. Collection is mandatory, automatic, and requires no effort from the individual beyond physical compliance. It happens at booking, takes seconds, and enters the person’s genetic information into a searchable national system. Removal is optional, manual, and requires the individual to discover the process, assemble documentation, and petition an agency that has no institutional stake in granting the request. The asymmetry is not a flaw in the system’s execution. It is a feature of its design. The system is built to capture efficiently and release reluctantly.

New bills continue to invoke expungement as the answer to constitutional objections. Kentucky’s HB 414 includes provisions for automatic removal upon acquittal or dismissal. Iowa’s HF 2624 provides that a sample is not required if the person’s DNA is already in the database and has not been expunged. These provisions may represent genuine improvements over the petition-only model. But they do not address the deeper problem. In a system that processes millions of arrests per year, even a small failure rate in expungement produces a large and growing population of unconvicted individuals whose genetic profiles remain permanently in law-enforcement databases. The safeguard works only if it works every time. Nothing in the system’s architecture ensures that it does.

What a Rights-Protective System Would Require

If the problems documented in this article are structural, the remedies must be structural too. Tinkering with expungement timelines or adding a procedural step to the booking process will not address the core issue, viz., the state has claimed the power to collect, store, search, and share the genetic identity of millions of people on the basis of an accusation that may never be proved. A system that respected constitutional rights as seriously as it pursues investigative efficiency would look fundamentally different from the one that exists.

Require a Conviction Before Permanent Inclusion

The most direct reform would be to limit permanent DNA database inclusion to individuals who have been convicted, reserving arrest-stage collection for temporary, case-specific use only. This would not eliminate DNA’s investigative value. It would restore the principle that the state must prove its case before imposing lasting consequences on a person’s genetic privacy. Profiles collected at booking could be compared against the existing forensic index for the duration of the pending case and then destroyed if no conviction results. The current system reverses this logic, retaining the profile by default and requiring the individual to seek its removal.

Mandate Automatic Expungement with Verified Destruction

For states that maintain arrest-stage collection, the minimum acceptable safeguard is automatic, state-initiated expungement that extends to both the digital profile and the biological sample. New York’s Assembly Bill A2441, introduced in the 2025–2026 session, offers one model. It would require automatic expungement of juvenile records, including DNA and other genetic material, upon termination of proceedings in favor of the respondent. The bill prohibits retention of juvenile DNA in any government database other than those used for parentage or juvenile offender proceedings. Similarly, Kentucky’s HB 414 provides for automatic expungement upon acquittal or dismissal. These approaches are preferable to petition-based models because they do not require the person to initiate their own protection. But automatic expungement must be paired with verification mechanisms that confirm both the profile’s removal from all database tiers and the physical destruction of the stored sample. Without verification, the “automatic” label becomes another paper safeguard.

Regulate Familial Searching
and Informal Collection

Montana’s House Bill 602, enacted in 2021, offers a model for regulating the investigative uses of genetic data. The law requires a warrant before law enforcement can use results from consumer DNA databases or familial searches in a criminal investigation. Extending a similar warrant requirement to familial searches conducted within CODIS would impose judicial oversight on a technique that currently operates at the discretion of individual states and laboratories. Maryland and the District of Columbia have banned familial searching outright. At a minimum, states that permit the practice should require judicial authorization and should prohibit its use for offenses below a specified severity threshold.

Informal DNA collection also needs a statutory framework. The street-level practices documented earlier in this article – officers asking for cheek swabs during field encounters, collecting DNA from minors without parental consent, and harvesting genetic material from discarded objects – currently operate in a legal vacuum. Legislation should require that any DNA collection outside a formal arrest be preceded by written, informed consent that clearly states the person’s right to refuse without consequence and that minors cannot consent without a parent or guardian present. New York’s Assembly Bill A6445, also introduced in the 2025–2026 session, addresses a related concern. It would restrict the inclusion and unrelated use of DNA collected from sexual-assault victims, recognizing that DNA provided for one purpose should not be repurposed for another without the individual’s knowledge.

Destroy Biological Samples After Profile Generation

Finally, the retention of biological samples after the STR profile has been generated should end. The government’s stated justification for DNA collection is identification and investigation through the CODIS profile. If the profile serves those purposes, there is no legitimate reason to retain a biological specimen containing the individual’s complete genetic blueprint. Mandatory destruction of the sample after profile generation would eliminate the possibility of future re-analysis using more advanced techniques and would close the gap between what the system claims to need and what it actually keeps. As long as the sample survives, the informational scope of the original collection can expand without any additional legal authorization.

The Presumption of Innocence in the Age of Genetic Policing

In January 2026, a man stood on a Minneapolis sidewalk and filmed federal immigration officers. He was tackled, handcuffed, and swabbed. No charges followed. NPR reported that in the past DNA specimens taken by federal immigration officers have been added to a national database maintained by the FBI, but it also reported that it was unclear where the samples acquired from protesters in recent months were ending up or how they were being used. In 2009, Lily Haskell was arrested at a peace rally in San Francisco. She was told to provide a DNA sample or face additional charges. No prosecution ever materialized. More than 15 years later, her genetic profile had still not been removed from the system.

Between those two events, the architecture documented in this article took shape. Thirty-one states, the District of Columbia, and the federal government now authorize the collection of DNA from people who have been arrested but not convicted. The FBI’s national index holds more than six million arrestee profiles. Those profiles are searched automatically and continuously against crime-scene evidence from across the country. Familial searching extends the investigative reach of each stored profile to the biological relatives of the person swabbed. And the expungement mechanisms that are supposed to protect the unconvicted fail often enough that large numbers of profiles persist in the database indefinitely.

The legal foundation for all of this rests on a single analogy: that a cheek swab is essentially the same as a fingerprint. The Supreme Court adopted that analogy in King. Legislatures have repeated it in every expansion debate since. But as this article has shown, the analogy does not survive contact with what the system actually does. A fingerprint identifies a person. A DNA profile in CODIS identifies a person, investigates whether they committed unrelated crimes, exposes their family members to genetic scrutiny, and persists as a permanent investigative resource regardless of whether the arrest that triggered collection led to anything at all. The distance between the analogy and the reality is the distance between a booking procedure and a surveillance infrastructure.

Justice Scalia warned in his King dissent that the decision contained no limiting principle. He predicted that the logic of the ruling would be extended beyond its stated boundaries. That prediction has been confirmed at every level. Collection has expanded from violent felonies to all felonies, from felonies to misdemeanors, from criminal suspects to immigration detainees, and from detainees to political protesters. At no point has a court, a legislature, or a federal agency articulated a clear stopping point. The system grows because growth is its default condition and because every new profile added to the database strengthens the institutional argument for adding more.

The presumption of innocence is supposed to mean that the state bears the burden of proving guilt before it imposes the consequences of a criminal accusation. Pre-conviction DNA collection inverts that principle at the biological level. It captures a person’s genetic identity on the strength of an arrest alone, stores it in a searchable national database, and in most jurisdictions leaves the individual to initiate removal through an expungement process. The question this system poses is not whether DNA evidence has value. It does. The question is whether a society committed to the rights of the accused can accept a framework in which millions of unconvicted people are subjected to lasting genetic surveillance because they were once arrested, and the state collected what it could.  

 

Sources: 28 C.F.R. pt. 28 (2026); 34 U.S.C. § 12592 (2026); ACLU of N. Cal., ACLU Lawsuit Challenges California’s Mandatory DNA Collection at Arrest (Nov. 15, 2009); ACLU of N. Cal., It’s Your DNA … Or Is It? ACLU Suit Aims to Keep Your Genetic Blueprint out of the Government’s Hands (Sept. 5, 2013); ACLU of S. Cal., Ninth Circuit Will Re-Hear Mandatory DNA Collection at Arrest Lawsuit (Sept. 19, 2012); Meg Anderson, ICE Officers Are Taking DNA Samples from Protesters They’ve Arrested, NPR (Mar. 18, 2026); Arnold Ventures, A New Commitment to Solving Serious Crimes (Oct. 9, 2025); Arnold Ventures, Arnold Ventures Welcomes New Legislation to Expand DNA Collection and Solve More Crimes in Missouri (last visited Mar. 2026); Ariz. Rev. Stat. § 13-610 (2026); Bureau of Just. Stats., Felony Defendants in Large Urban Counties, cited in Crime in America (last visited Mar. 2026); Colo. Bureau of Investigation, DNA Database Expungement (last visited Mar. 2026); DNA-Sample Collection From Immigration Detainees, 85 Fed. Reg. 13,484 (Mar. 9, 2020); Elec. Privacy Info. Ctr., Maryland v. King (last visited Mar. 2026); Emma Kenny-Pessia, Ditching “DNA on Demand”: A Harms-Centered Approach to Safeguarding Privacy Interests Against DNA Collection and Use by Law Enforcement, 101 Wash. U. L. Rev. 497 (2023); Erin Murphy & Jun H. Tong, The Racial Composition of Forensic DNA Databases, 108 Calif. L. Rev. 1847 (2020); FBI, CODIS and NDIS Fact Sheet (last visited Mar. 2026); FBI, CODIS-NDIS Statistics (Nov. 2025) (last visited Mar. 2026); FBI, DNA Fingerprint Act of 2005 Expungement Policy (last visited Mar. 2026); Fed. Jud. Ctr., Law Enforcement Databases: Limited Genetic Information and Varying Procedures for Use (last visited Mar. 2026); Greg Brower, Maryland v. King: Possibly the Most Important Criminal Procedure Case in Decades, Fed. Soc’y Rev. (last visited Mar. 2026); Idaho State Police, Qualifying Misdemeanors and Owed DNA Collection: A Systematic Review of State-Level Statutes (Apr. 2025); 730 Ill. Comp. Stat. 5/5-4-3 (2026); Iowa Gen. Assembly, Bill History for H.S.B. 571/H.F. 2624 (2026); Iowa Gen. Assembly, Fiscal Note for H.F. 2624 (2026); House Passes Bill Expanding DNA Collection by Law Enforcement, Iowa Pub. Radio (Mar. 11, 2026); KSHB, Bill in Missouri Legislature Would Expand DNA Collection to More Felony Arrests (Mar. 2026); Ky. Legis. Rsch. Comm’n, H.B. 414, 2026 Reg. Sess.; Lauren Kirchner, DNA Dragnet: In Some Cities, Police Go From Stop-and-Frisk to Stop-and-Spit, ProPublica (Sept. 12, 2016); Lyle Denniston, Opinion Recap: Solving “Cold Cases” Made Easier, SCOTUSblog (June 3, 2013); Missouri Senate, Bill Information for S.B. 1458 (2026); Missouri Senate, Fiscal Note for S.B. 1458, LR No. 6299S.01I (Mar. 10, 2026); Mont. H.B. 602, 67th Leg., Reg. Sess. (2021); N.C. Crim. Just. Info. Network, What States Collect DNA on Arrest and Who Is Responsible for Expunging DNA if Applicable? (Oct. 2024); Nat’l Conf. of State Legislatures, DNA Collection After Arrest Laws (last visited Mar. 2026); Nat’l Hum. Genome Rsch. Inst., Investigative Genomics (last visited Mar. 2026); Nat’l Inst. of Just., Debating DNA Collection (last visited Mar. 2026); Nat’l Inst. of Just., DNA Evidence: Basics of Analyzing (Aug. 8, 2012); Nat’l Inst. of Just., DNA Sample Collection from Arrestees (Dec. 6, 2012); Nat’l Inst. of Just., Post-PCAST Court Decisions Assessing the Admissibility of Forensic Science Evidence (last visited Mar. 2026); Nat’l Inst. of Just., Understanding Familial DNA Searching, Grant No. 251080 (last visited Mar. 2026); Nat’l Inst. of Just., When DNA Samples Are Complicated: Calculating Variation in Mixed Samples Interpretation (last visited Mar. 2026); Nat’l Registry of Exonerations, About (last visited Mar. 2026); N.Y. Assemb. A2441, 2025-2026 Leg., Reg. Sess. (2025); N.Y. Assemb. A6445, 2025-2026 Leg., Reg. Sess. (2025); Norman L. Reimer, The Scalia Dissent in Maryland v. King: Exposing a Contrived Rationale Today and a Dangerous Precedent for Tomorrow, Fed. Soc’y (last visited Mar. 2026); President’s Council of Advisors on Sci. & Tech., An Addendum to the PCAST Report on Forensic Science in Criminal Courts (Jan. 2017); President’s Council of Advisors on Sci. & Tech., Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature-Comparison Methods (Sept. 2016); Ann Givens & Robert Lewis, In New York City, Gun Cases Fuel Growing, Unregulated DNA Database, The Trace & WNYC (Sept. 4, 2017); Collecting Migrant Children’s DNA: A Troubling, but Predictable, Trend in the U.S., Hastings Ctr. Bioethics Forum (July 7, 2025); DNA Upon Arrest: Solving Cold Cases or Presuming Guilt?, Stateline (Jan. 12, 2017); Note, The Fourth Amendment’s Failure to Protect Against Mass DNA Collection at the Border, 75 Duke L.J. (2025).

Subscribe today

Already a subscriber? Login