The Biden administration's successor announced a sweeping review of more than 55 million people who hold valid U.S. visas and is deploying artificial intelligence to scan records and flag potential violations, the State Department and former officials said, in what critics call an unprecedented immigration sweep.
The State Department has described the process as "continuous vetting" designed to identify visa overstays, criminal activity or national security risks. Officials said the effort will include automated cross-checks of government databases, scrutiny of social media accounts and reviews of immigration records from other countries.
Former State Department employees and immigration experts said the announcement appears aimed at a broad public showing of enforcement that could prompt some visa holders who believe they are ineligible to leave voluntarily. A former department employee described the approach to the Daily Mail as a kind of psychological operation intended to encourage self-deportation rather than to process all 55 million records individually.
"They don't need to scrub 55 million. They just need to say they are casting the net as extensively as possible," the former employee said. The official also said the review will prioritize certain countries but that explicit targeting would not be publicly acknowledged.
Administration officials and some current department staff cautioned that the public figure of more than 55 million represents the total universe of visa holders and that practical enforcement is expected to focus on smaller, higher-priority groups. The State Department has not released a public timeline or a full description of the algorithms and data sources that will be used.
Immigration policy specialists said the plan raises practical and legal questions about accuracy, transparency and due process. Julia Gelatt, associate director of the U.S. Immigration Policy Program at the Migration Policy Institute, said different government databases have gaps and errors that could lead to erroneous visa revocations.
"Different government databases are speaking to each other looking for matches, but there are concerns some have incomplete information," Gelatt said. She pointed to cases in which arrests or interactions with law enforcement have remained in records without context, potentially triggering automated flags even when individuals were not charged or were found innocent.
Technology analysts warned that AI systems optimized for speed risk producing false positives. Rob Enderle, president and principal analyst at the Enderle Group, said the platforms currently used by some government offices prioritize productivity over exhaustive human review, increasing the likelihood of mistakes.
Concerns about errors are not theoretical. In April, officials reversed a visa termination for a Japanese student at Brigham Young University after what attorneys described as an AI-related mistake tied to a minor traffic citation and a fishing violation. Separately, a Turkish student at Tufts University was arrested in March after an F-1 visa was revoked and the student was transferred to Immigration and Customs Enforcement custody, prompting criticism from lawmakers and civil rights groups who said political factors were at play.
Since January, the administration reports roughly 6,000 student visas have been revoked, with about 4,000 attributed to law violations, according to the State Department. Department of Homeland Security statistics cited in recent reporting note nearly 13 million green-card holders and about 4 million people in the U.S. on temporary visas last year.
Former and current officials also highlighted logistical constraints. The State Department reduced staff in consular affairs by roughly 20 percent, officials said, which supporters of the AI-driven approach argue makes automated tools necessary to process large data sets. Critics say that staffing cuts make rigorous human review less likely and could compound the risk of wrongful actions.
Advocates and policy researchers called for more transparency about the vetting procedures, the data sources to be used and safeguards for people flagged by automated systems. Gelatt said the government should publish details on testing and error rates before expanding automated vetting and ensure meaningful avenues to contest results.
A State Department official cited by Fox News said revocations under the current administration have targeted individuals who have broken the law or expressed support for terrorism. The department has not publicly released the procedures for the new continuous-vetting program or provided a comprehensive accounting of the algorithms and third-party data providers that will be used.
Immigration experts and civil liberties groups said oversight, public reporting and statutory protections will be essential if AI is expanded for immigration enforcement, noting the balance between national security objectives and the risk of irreversible consequences for lawful visa holders. The State Department and Department of Homeland Security did not respond to requests for further comment beyond previously issued statements.