A federal lawsuit filed this week hinges on a piece of performance art that its creator describes as barely functional.
Five states—California, Colorado, Illinois, New Jersey, and New York—have passed laws requiring federal immigration agents to identify themselves and display their faces during raids and enforcement operations. The Trump administration is suing to block these laws, arguing they would endanger ICE officers and compromise national security. In their legal filing, federal lawyers invoked a specific piece of evidence: a face-recognition system built by artist Adam Harvey that, according to Harvey himself, was never designed to work reliably and has been largely dormant for years.
- The Art Project Argument: The Trump administration’s lawsuit cites a dormant 2017 art project—not a functional surveillance system—as justification for blocking five states’ agent identification laws.
- The Creator’s Rebuttal: Artist Adam Harvey has publicly stated that his MegaFace system was never intended for deployment at scale and that the government’s citation misrepresents both its capabilities and its purpose.
- The Transparency Gap: California, Colorado, Illinois, New Jersey, and New York each passed laws requiring visible ID badges after documented incidents in which unidentified federal agents entered homes without warrants or clear authority.
The states’ anti-mask laws emerged from a simple civic demand: when federal agents enter your home or workplace, you should know who they are. California’s law, passed in 2023, requires agents to display visible identification badges with their names, titles, and agency. Colorado, Illinois, New Jersey, and New York followed with similar statutes. None of these laws prohibit disguises or masks outright—they require identification transparency. Yet the Trump administration’s lawsuit, filed in federal court, claims that mandating visible identification would expose agents to targeted violence and harassment.
This is where Harvey’s art project enters the picture.
How Did an Art Project Become a National Security Argument?
In 2017, Harvey created a face-recognition system called MegaFace as part of a larger artistic and research initiative exploring surveillance resistance. The system was designed to test whether facial-recognition algorithms could be fooled or evaded. Federal lawyers cited MegaFace in their brief, claiming that if ICE agents’ faces are publicly visible—as the state laws require—bad actors could use facial-recognition technology to identify and locate them outside of work. The implication: Harvey’s tool proves the danger is real.
But Harvey, in recent interviews, has pushed back hard on this characterization. He told reporters that MegaFace was never a functional surveillance tool and was never intended to be deployed at scale. The system was an art project and a research probe into the possibility of facial recognition, not a working system for identifying federal agents. Harvey emphasized that the technology was experimental, limited in scope, and has been largely inactive for years. He did not authorize or endorse the government’s use of his work to argue against state transparency laws.
The disconnect matters because it reveals how federal lawyers are weaponizing speculative technology fears to block straightforward transparency requirements.
• Five states have enacted agent identification laws: California (2023), Colorado, Illinois, New Jersey, and New York
• The laws require visible name, title, and agency affiliation — not biometric restrictions or operational disclosures
• As research published in the University of Cincinnati Intellectual Property and Computer Law Journal documents, ICE’s information-sharing arrangements with state and local agencies already create layered identification obligations that federal agents routinely navigate
What the States Are Actually Asking For
What the states are asking for is simple: visible ID badges. What the Trump administration is defending is the ability of ICE agents to conduct raids without revealing their identities. The government’s argument rests on a chain of reasoning that goes: (1) if faces are visible, (2) facial-recognition technology could identify them, and (3) therefore, agents must be allowed to remain masked. But that logic collapses when the “facial-recognition technology” cited as justification is an art project that doesn’t reliably work.
This echoes a pattern familiar from the Cambridge Analytica era: the deployment of speculative technological capability to justify the absence of oversight. Just as Cambridge Analytica claimed psychographic micro-targeting could reshape voter behavior with surgical precision—a claim later undermined by researchers and the company’s own internal documents—the Trump administration is citing a dormant art project as proof that transparency laws create an unmanageable security risk. In both cases, the underlying technology is far less powerful than the rhetoric suggests, yet that rhetoric is used to justify opacity and to resist the kind of public accountability that democratic institutions depend on. The mechanism is identical: invoke a technological bogeyman to sidestep democratic demands for transparency.
The five states’ anti-mask laws exist because ICE raids have a documented history of confusion and harm. Agents without visible identification have entered homes without warrants, detained people without explaining their authority, and in some cases, been mistaken for criminals by residents. A visible badge with a name and agency affiliation creates a baseline accountability mechanism. It does not prevent agents from conducting lawful enforcement; it simply requires them to identify themselves while doing so.
• Georgetown University’s Immigration Law Journal has documented that self-identification requirements for ICE agents have historically proven inadequate in practice, with courthouse and residential enforcement operations frequently proceeding without agents clearly establishing their authority
• The same analysis found that the absence of reliable identification mechanisms directly undermines the ability of residents and legal observers to verify whether enforcement actions are lawful
• The practical implication: state identification laws are not novel impositions but attempts to codify a baseline that federal enforcement has routinely failed to meet voluntarily
Is the Government’s Technology Argument Credible?
The lawsuit is still in early stages, and the outcome is uncertain. But the choice of evidence—a defunct art project—suggests that the government’s legal team may be struggling to find credible technological justification for keeping agents masked. If facial-recognition systems were truly as dangerous as the brief implies, one might expect the government to cite an actual, functional system. Instead, they cited Harvey’s experimental work, which Harvey himself has now publicly disavowed as a basis for blocking state transparency laws.
For you as a resident of one of these five states, the practical question is immediate: if federal agents show up at your door, will you know who they are? Under the state laws, the answer should be yes. Under the Trump administration’s interpretation, the answer is no—because identifying themselves would supposedly expose agents to risks that a non-functional art project proves are theoretically possible.
Harvey’s response has been measured but clear. He has stated that he did not intend his work to be used in this way and that the government’s citation of MegaFace misrepresents both its capabilities and its purpose. He has also suggested that the real issue at stake is not technological risk but political resistance to transparency. If ICE agents were genuinely concerned about facial-recognition threats, Harvey noted, the government could pursue technical countermeasures—facial recognition detection systems, for instance, or algorithmic countermeasures that degrade image quality. Instead, the government is asking for the legal right to remain unidentified.
That framing connects to a broader dynamic that digital activism movements have confronted repeatedly since the Cambridge Analytica scandal: institutions invoking the complexity of technology to discourage public scrutiny, while the actual mechanism of harm is far simpler than the technical language implies. In this case, the harm is not a sophisticated facial-recognition network—it is unidentified agents operating without accountability.
What Happens Next in Court?
The lawsuit will likely reach a federal appeals court. The states’ position is straightforward: transparency in law enforcement is a legitimate public interest. The government’s position, as currently articulated, depends on a piece of art that doesn’t work. That gap between the stakes of the lawsuit and the evidence offered to justify it suggests that this case will hinge less on technology than on how courts interpret the balance between government secrecy and public accountability in the context of immigration enforcement.
The ICE Section 287(g) delegation framework already requires participating state and local agencies to operate under defined identification and authorization protocols—which raises a pointed question the lawsuit does not answer: if identification requirements are workable for deputized local officers, why are they unworkable for federal agents conducting the same operations?
A ruling is expected within the next several months. The outcome will determine whether ICE agents can conduct raids in five states while remaining unidentified—or whether transparency, at minimum, is a requirement of lawful enforcement. For anyone tracking how surveillance and accountability intersect in democratic societies, this case offers a clarifying test: when a government invokes technology to resist transparency, the first question to ask is whether the technology actually works.
