Here is a scenario that should concern anyone who could end up in a police lineup, physical or digital. A camera captures your face. Software scans it against a database, returns a name, and that name becomes the reason officers show up at your door. You are arrested, charged, and handed a discovery file. Nowhere in that file does the algorithm appear. As far as the paperwork is concerned, a detective simply developed a lead.
That gap between how you were actually identified and what your defense attorney is allowed to see is the fight now playing out in state legislatures and appellate courts. In mid-2026, two states moved to pull police facial recognition out of the shadows of the case file, and a federal bill is pushing to require a warrant before the search happens at all. For defendants, the through-line is simple: if an algorithm drove your case, will you ever be told, and can you challenge a match that officially is not evidence?
The Maryland model: the law advocates call the strongest in the country
Maryland's facial recognition statute, passed as HB 338 and SB 182, took effect on October 1, 2024. StateScoop and the Security Industry Association both describe it as the strongest state law-enforcement facial recognition law in the country, and the reason is that it targets the exact places where a match can quietly shape a prosecution.
Four provisions matter most for anyone facing charges:
-
No bare match as evidence. A facial recognition "potential match" result cannot be used as evidence against a defendant. The software output is treated strictly as an investigative lead, not proof of identity.
-
Mandatory human review. Trained, independent human review is required before an FR lead can be acted on for further investigation. An officer cannot take a raw software hit and run with it.
-
No real-time identification. The law bars live or real-time identification, closing off the continuous-surveillance use of the technology.
-
Discovery disclosure. When facial recognition led to further investigative action, prosecutors must disclose that use in discovery, including the names of each FR system used and the databases that were searched.
That last point is the one that reaches your defense directly. A match that is not admissible as evidence can still be the hidden engine of the entire case. Maryland's answer is to require that the engine be named on the record, so your attorney knows to probe it.
The June 30, 2026 training deadline
Maryland's law does not just restrict facial recognition, it conditions its use on training. The Department of Public Safety and Correctional Services is required to develop a facial recognition training program covering FR use, cultural diversity, and implicit bias, with a completion deadline of June 30, 2026.
The deadline operationalizes the human-review requirement. Before an FR lead can properly be acted on for further investigation, the officers involved are meant to have gone through training built around the known failure modes of the technology, including bias in how algorithms perform across different groups. For a defense attorney, the training records and whether they were completed become a concrete line of inquiry about how an identification was handled.
Johnson v. State: when non-disclosure becomes a fair-trial violation
Statutes set the rules. Courts decide what happens when the rules are broken. In August 2025, the Appellate Court of Maryland delivered the answer defendants were waiting for in Johnson v. State.
According to Public Justice, which tracked the ruling, the State disclosed its use of facial recognition only days before trial. The appellate court held that this late disclosure deprived the defendant of a meaningful opportunity to challenge the AI-generated identification and violated the defendant's fair-trial and Brady obligations. The court reversed the conviction.
The significance is in how the ruling converges with the statute. The law says FR use must be disclosed in discovery. Johnson confirms that dumping that disclosure on the eve of trial is not enough, because a defendant needs real time to investigate and contest the identification. Together they establish that facial recognition is not a detail prosecutors can bury or reveal at the last minute. It is disclosable, and disclosing it late can cost the State the conviction.
Virginia's July 1 change: a pointed ban on real-time tracking
Virginia's approach is narrower but sharp. The revised Va. Code section 15.2-1723.2 takes effect July 1, 2026. Its headline prohibition bars local law-enforcement agencies from using facial recognition to track the movements of an identified individual in a public space in real time.
The statute also builds in accuracy and accountability guardrails. Facial recognition algorithms must score at least 98% true positives in the NIST Face Recognition Vendor Test, and agencies are barred from building databases out of live video feeds. The penalties fall on operators: a first violation is a Class 3 misdemeanor, and a second is a Class 1 misdemeanor plus termination.
Where Maryland attacks the evidence-and-disclosure problem, Virginia attacks a specific surveillance capability, following an identified person through public space as they move. The NIST accuracy floor is a defense-relevant detail on its own, because it sets a measurable standard that a challenged identification can be tested against.
The federal wildcard: a warrant requirement and a private right of action
At the federal level, the effort is earlier and less certain. On April 23, 2026, Representatives Thomas Massie of Kentucky and Lauren Boebert of Colorado introduced the Surveillance Accountability Act, H.R. 8470. The bill would require a warrant based on probable cause before facial recognition and other biometric searches, and it would create a private right of action for Fourth Amendment violations.
That combination is what makes the bill notable for defendants. A warrant requirement would move facial recognition searches from an informal investigative step to one that a judge must authorize in advance. A private right of action would give people a direct path to sue over unlawful biometric searches rather than relying solely on suppression inside a criminal case.
The odds, however, are uncertain. The bill sits in the House Judiciary Committee, and its sponsors reflect a libertarian rather than broadly bipartisan coalition. It is a signal of where the debate is heading more than a settled change in the law.
The defense-rights takeaway
Strip away the state-by-state detail and the same question runs through all of it. If you were identified by an algorithm, will you be told, and can you challenge a match that is not itself evidence but drove the whole investigation?
Maryland's answer is that the match cannot stand in as evidence, must survive trained human review, and must be disclosed in discovery, with Johnson v. State making late disclosure a fair-trial violation. Virginia's answer is to cut off real-time tracking of identified people and hold operators personally accountable. The federal proposal would add a warrant at the front end. If you are facing charges and suspect facial recognition played a role in identifying you, the disclosure question is not a technicality. It is the difference between challenging how you were identified and never knowing it happened.
Related reading
-
[The .05 Blood-Alcohol Line Is Spreading: Washington Eyed a July Cutover, and Utah Is No Longer the Only Bill on the Table](/article/05-bac-limit-spreading-washington-utah)
-
[The Meth 'Purity Loophole' That Doubles Federal Sentences Survives: The Reform the Sentencing Commission Proposed, Then Quietly Shelved](/article/meth-actual-vs-mixture-purity-sentencing-reform-shelved-2026)
-
[Half of Sober Drivers Fail the Roadside Marijuana Test: Why Courts Are Starting to Throw Out Cannabis DUI Field Evidence](/article/sober-drivers-fail-roadside-marijuana-test-cannabis-dui-field-evidence)
Sources
-
Maryland lawmakers approve 'strongest' facial recognition rules for law enforcement yet, StateScoop
-
H.R.8470, Surveillance Accountability Act, 119th Congress, Congress.gov
-
Reps. Massie and Boebert Introduce the Surveillance Accountability Act, Rep. Thomas Massie
Note: This article contains AI-assisted content and has been reviewed by our editorial team.
IMPORTANT NOTICE: GavelDaily.com is an advertising and information service, not a law firm. Submitting this form does not create an attorney-client relationship. The information you provide will be shared with licensed criminal defense attorneys in your area who may contact you about your inquiry. All submissions are treated as confidential by our service, but attorney-client privilege does not attach until you formally retain an attorney. Any attorney you connect with through this service is independently responsible for their legal advice and representation. Free consultations are subject to individual attorney availability. If you are facing an immediate legal emergency, contact your local public defender's office or call 911.