In Focus
10 Arrested After Armed Ambush on Texas ICE Detention Facility
Launch Slideshow 2 PHOTOS
HOUSTON — Federal enforcement records show that between February 2025 and February 2026, agents in Texas took more than 38,000 people into immigration custody who had neither a criminal conviction nor an open criminal case — outnumbering both people with convictions and people facing pending charges combined, according to an analysis published by the Houston Chronicle and summarized by Texas Standard. The finding sits awkwardly next to the administration's original pitch for its deportation campaign, and it falls heaviest on Latino households in Houston and across the state, since Latino immigrants make up the large majority of people caught up in these sweeps.
A Pledge to Target "Dangerous Offenders" Meets a Different Reality
Trump campaigned in 2024 on removing what he called the country's most dangerous unauthorized immigrants, a promise his team later narrowed to mean people convicted of or charged with serious crimes. Records obtained through a public-records request by the Deportation Data Project and reviewed by Chronicle reporter Julián Aguilar point to a different pattern in practice: enforcement pursued that stated target only in the opening months of the term before increasingly turning toward people whose only violation involved how they entered the country or an expired visa — matters typically handled through civil immigration proceedings rather than criminal court.
The Department of Homeland Security disputes this framing. It argues that close to 70% of arrests nationwide involve people it classifies as "criminal illegal aliens," and it has suggested that some people counted as clean-record cases may carry offenses from other countries that never surface in U.S. databases. Aguilar said that the agency hasn't backed up that claim with the kind of documentation it usually releases after major criminal sweeps, when it typically publicizes mugshots and case details. Instead, he said, officials tend to "sort of quibble with the numbers and say that they're being cherry-picked."
What DHS Actually Said About the Houston Numbers
Local data mirrors the statewide pattern. Inside the Houston field office's territory — dozens of counties running from the Gulf Coast up toward the Waco area — only 13% of the roughly 1,077 people arrested in January 2025 had no criminal charge or conviction. By late June, monthly arrests had risen 73% to nearly 1,900, and the noncriminal share had grown to 37%, according to Axios Houston's review of Deportation Data Project figures — part of a national swing from roughly a fifth of daily arrests in early May to close to half by early June, after the administration tripled ICE's arrest quota that month.
It's worth being precise about how DHS responded to this data, because the agency's own words are easy to misread. DHS assistant secretary Tricia McLaughlin didn't concede that ICE was avoiding criminal cases — she pushed back on the opposite claim, telling Axios that the notion "ICE is not targeting criminal illegal aliens" is a false narrative pushed by the media. She added that about 70% of arrests involved people with convictions or pending charges, without providing supporting detail for that number.
A Statewide Surge Rooted in Local Jails
Widen the lens to the full state and the same pattern shows up over a longer stretch. A Texas Tribune review comparing the final year and a half of the Biden administration with Trump's first six months back in office found that Texas's average daily ICE arrests climbed from 85 to 176, in a state home to an estimated 2 million residents without legal status. More than half of those arrests began at local jails, a pipeline widened by a Texas law now requiring every county that operates a jail to enter a formal partnership with ICE.
A separate count from the Prison Policy Initiative put Texas's arrest rate at roughly 58 people per 100,000 residents in the four months before the administration's quota increase took effect in May 2025 — nearly doubling to about 110 per 100,000 in the four months that followed.
Latino Neighborhoods Feel the Brunt
Because Latino immigrants make up the overwhelming majority of people targeted in these operations, researchers argue that the shift away from criminal cases functions, in effect, as a policy that reshapes Latino neighborhoods specifically. A UCLA Luskin School analysis found that the share of Latino arrestees who met the government's own bar for a genuine public-safety threat fell from 28% under the Biden administration to 12% under Trump. By early 2026, only about one in ten arrests involved someone in that serious-threat category, compared with roughly one in five in early 2024.
Local advocates describe the same trend in blunter terms. After a roughly 1,500-arrest, 10-day operation across Houston last fall that ICE promoted as a public-safety success, Sergio Lira of Greater Houston LULAC told Click2Houston that "this is harming our community in the long run."
Showing Up for a Check-In — and Getting Detained
Immigration attorneys point to one group in particular swelling the noncriminal numbers: asylum seekers who voluntarily attend scheduled ICE check-ins tied to claims filed years earlier, in some cases dating back to the Biden administration or even Trump's first term. Aguilar has noted that people who keep returning to a federal office on their own would seem, on its face, unlikely to be hiding something serious — a pattern that has led advocates to question whether these arrests reflect pressure to hit quotas more than any broader public-safety strategy.
What Comes Next
The data project behind these figures continues to release new updates, and DHS and outside researchers are likely to keep disputing how "criminal" should be counted in enforcement statistics. For now, the space between the administration's original target — dangerous offenders — and the people actually showing up in arrest logs remains the central tension in Texas's immigration enforcement debate, with Houston's Latino communities watching the numbers most closely.
