ICE’s FY2026 data show a June surge in arrests and higher detention, but “criminals first” doesn’t match the custody mix

ICE’s FY2026 enforcement picture, as revealed in July, is straightforward on one point and messier on several others: arrests jumped sharply in June, detention rose by early July, and the administration’s preferred “criminals first” narrative does not fully explain who is in custody.

The basic measurement frame matters. ICE says its enforcement data are published in arrears and may later be corrected. OHSS says its monthly tables are standardized from DHS operational reports and are the official ledger for monthly arrests, detention book-ins, book-outs, and removals/returns. Here, book-ins means people newly entering ICE custody in that month; book-outs are people leaving custody, whether by release, transfer, or the end of detention; and removals/returns track people ICE sends out of the country or processes out of custody. That means the July release is usable, but provisional; it is a versioned snapshot, not the last word.

Within that framework, the headline movement is the June arrest surge. Secondary reporting put June arrests/book-ins at 43,138, which multiple outlets described as a record or near-record monthly total. We do not directly extract the underlying June row here, so the exact monthly total should be treated as reported rather than table-verified in this pass. But the direction is not in dispute: ICE enforcement rose sharply.

Detention rose too. USA Today reported 65,765 detainees as of July 11, and earlier reporting noted that the detained population was still below January’s peak but higher than in April. What is not yet proven from the raw tables in this pass is whether removals accelerated enough to keep pace. The secondary reporting leans toward no: USA Today and GovExec both suggested removals were roughly flat even as detention climbed, implying accumulation in custody rather than faster throughput alone. That is a plausible reading, not a fully extracted finding.

The public argument over the surge is really about who is being held. ICE’s own statistics divide detainees into three buckets: people with convictions, people with pending criminal charges, and people with no known criminal history. In ICE’s data, that third bucket means no known convictions or pending charges in the agency’s records; it can still include repeat reentrants, immigration fugitives, and other people who have broken immigration law. So “no known criminal history” does not mean innocence, and it does not mean there was no enforcement basis. But it does complicate any attempt to say the crackdown is only about criminal offenders.

That complication is visible in the reporting. USA Today said the no-criminal-history group remained the largest share of detention, citing a 37% figure. GovExec reported a similar picture, saying fewer than one-third of detainees had criminal convictions and about 40% had only non-criminal immigration violations. Those numbers do not refute the existence of criminal detainees; they do undercut a simple story that ICE is mainly holding people with violent or serious records.

The administration’s preferred version is still clear enough to state fairly. DHS says the crackdown is focused on people with criminal records, and Reuters quoted a DHS spokesperson saying the administration is delivering on Trump’s promise to arrest and deport those people. That is the strongest form of the government’s case. The problem is that the detention mix, as reported, is broader than the rhetoric implies.

The legal constraint is just as important. Reuters reported on July 2 that a 5th Circuit panel said the government must provide bond hearings after 90 days of detention, and that DHS said it disagreed and remained confident in its mandatory-detention position. Reuters also reported in February that courts had ruled more than 4,400 times that ICE was detaining people unlawfully. Together, those reports show that the detention push is being actively contested in court, which matters for how durable any population increase will be.

So the best reading of the July data is not a triumphal crackdown on criminals, and not a proof that the whole detention increase is a backlog effect either. It is a verified June surge in arrests/book-ins, a higher detained population by early July, a still-large noncriminal share in custody, and a legal environment that may keep limiting how far the detention expansion can run.

FY2025–FY2026 monthly ICE enforcement counts (arrests/book-ins, detention population, removals/returns, bond releases)

Does June represent a true step-change in enforcement operations, and how does it compare with detention growth and removal/bond release trends in the months before and after? Monthly ICE initial book-ins to detention, average daily detention population, book-outs to remove, and bond releases for FY2025 through June 2026, from ICE’s congressionally mandated detention statistics (FY25 year-end and FY26 YTD posted 20 July 2026). June 2026 book-ins hit 43,138 — the highest month in this window and a clear jump from the ~32–34k range in February–May — while average daily detention stayed near 59k that month (after peaking above 71k in January) and book-outs to remove did not rise with the June intake surge; bond releases stayed near ~2,000. Publisher: U.S. Immigration and Customs Enforcement (ICE), Detention Management — https://www.ice.gov/detain/detention-management (files FY25_detentionStats09242025.xlsx and FY26_detentionStats07202026.xlsx). OHSS Immigration Enforcement and Legal Processes Monthly Tables remain delayed under review and do not yet publish FY2025–FY2026 monthly rows. ICE Removals are shown only as fiscal-year totals in these workbooks (319,980 in FY2025; 356,389 YTD in FY2026), so the monthly removals line is the official “Release to Remove” book-out series, not a full monthly removals/returns count. Values may be revised; July 2026 is incomplete in the source and is omitted. — AI-assisted analytic, built only from real cited or sourced data. Source: The Guardian, USA Today, Government Executive / Stateline. As of 2026-07-21.