How does ICE classify 'criminal' in its detention statistics and how has that changed in 2025?

Checked on February 6, 2026
Disclaimer: Factually can make mistakes. Please verify important information or breaking news. Learn more.

Executive summary

ICE counts “criminality” in its published detention tables using discrete categories—convicted criminal, charged/pending criminal case, and non‑criminal (which can include repeat re‑entries, immigration fugitives, or those with no recorded criminal charges)—a schema reflected in ICE’s biweekly “Currently Detained by Criminality” reporting . In 2025 ICE’s reported population shifted sharply: arrests and detention growth were overwhelmingly driven by people ICE classifies as non‑criminal or without convictions, reflecting both operational changes (more street arrests and transfers) and policy moves that broadened who is detained .

1. How ICE defines “criminal” in its public statistics

ICE’s public statistics separate detained people into categories based on conviction status and pending charges, and explicitly identify a third “non‑criminal” grouping that covers people without known criminal convictions or pending criminal charges as well as some immigration‑specific categories like repeat re‑entrants, immigration fugitives with final orders, and foreign fugitives wanted for crimes abroad . ICE’s Detention Management page frames detention as civil immigration custody but notes mandatory detention for certain categories set by statute, and ICE says it uses custody determinations to assess public‑safety or flight risk . The Guardian and ICE’s own tables use the “Currently Detained by Criminality” construct as the source for media and research breakdowns .

2. What the 2025 data show about who is labeled “criminal”

Multiple independent data aggregators and research groups found that by mid‑ and late‑2025 a large majority of people in ICE custody had no criminal conviction recorded in ICE’s own categories: analyses and point‑in‑time tallies report figures around 71–74 percent of detainees with no criminal conviction as of late 2025 and January 2026 . Researchers who parsed ICE’s biweekly spreadsheets found that virtually all of the increase in detention counts between late September 2025 and early January 2026 was attributable to people without convictions—one analyst concluded 92 percent of detention growth was among those ICE classifies as non‑criminal .

3. Why those numbers moved so fast in 2025

The shift reflects operational changes—ICE substantially increased street/community arrests, expanded at‑large apprehensions, and more often booked people who previously might only have been transferred from local jails—rather than a proportional rise in people with serious criminal convictions [1]. Reporting and advocacy groups document a dramatic rise in interior arrests of people without criminal records and note ICE’s increased reliance on transfers and temporary facilities to hold the larger population . ICE’s own statistics also show a growing share of detainees arriving from CBP and field arrests rather than solely from local jails .

4. Competing narratives and institutional agendas

The administration and DHS emphasize public‑safety rationales and statutory detention obligations when defending the expansion of detention , while watchdogs, legal advocates, and independent data analysts portray the trend as a policy choice to detain non‑criminal immigrants en masse to drive deportations . Leaked ICE data used by the Cato Institute and reporting groups further undercut political claims that detention growth targeted only “criminal aliens,” showing a high share of deportations and bookings involving people with no convictions .

5. Data caveats and limits on what can be concluded

ICE datasets have known limitations—biweekly point‑in‑time snapshots, inconsistent releases after a 2025 shutdown, missing identifiers that complicate tracking individuals across locales, and the exclusion of some DHS enforcement actions or federal criminal cases—so precise attribution of every arrest or the full scope of changes is constrained by data gaps [1]. Analysts nonetheless converge on the qualitative finding that 2025 enforcement and detention policy produced rapid growth concentrated among those ICE labels as non‑criminal .

Want to dive deeper?
How does ICE's 'Currently Detained by Criminality' table categorize repeat re‑entrants and immigration fugitives?
What legal memos or policy directives in 2025 changed ICE mandatory‑detention criteria and interior enforcement tactics?
How do local jails and law enforcement partnerships affect ICE's ability to classify and detain people as 'criminal' versus 'non‑criminal'?