What percentage of ice agents have a criminal background?
Executive summary
The sources provided do not contain any data on the criminal background of ICE employees or agents, so the direct question—what percentage of ICE agents have a criminal background—cannot be answered from this reporting; there is no empirical claim in these documents about the share of ICE staff with criminal records (p1_s1–[4]2). What the reporting does document in detail is the criminal-conviction status of people detained or arrested by ICE: multiple independent analyses and ICE-related datasets show a large and growing share of detainees lack criminal convictions, a fact that is sometimes conflated in public debate with questions about ICE personnel themselves [1] [2] [3].
1. The user’s question vs. the available evidence
Careful parsing shows two distinct questions have been conflated: one about the criminal histories of ICE agents (personnel) and another about the criminal-conviction status of people ICE arrests and detains; the supplied documents answer the latter but are silent on the former, so any attempt to state a percentage of agents with criminal records would be unsupported by these sources [4] [5]. The datasets and reporting assembled here—TRAC, Cato, Stateline, MinnPost, and ICE’s own stats—speak repeatedly about detainees’ records and enrollment in detention, not about background checks or disciplinary records for ICE employees [1] [3] [6] [2] [5].
2. What the reporting actually says about criminal records — of detainees, not agents
Across the sources, the consistent finding is that a majority of people in ICE custody in recent months had no criminal convictions on file: TRAC reported roughly 73.6% of people in ICE detention had no criminal conviction as of November 30, 2025 (48,377 of 65,735) [1], MinnPost summarized ICE’s own figures showing about 71% with no convictions in 2025 [2], and the Cato analysis similarly found that a large share of arrests involved people without criminal convictions, with only about 5% of detainees having violent convictions in one dataset [3] [7]. Stateline’s analysis of a narrow late‑2025 window even found the share with convictions fell to single digits (noting a 3% figure for a particular period), underscoring how point-in-time and rolling-window measures can vary sharply [6]. ICE’s public materials emphasize the mix of people detained — from those with serious convictions to those with no convictions or only minor offenses — but do not provide a staff‑background metric [5].
3. Why the distinction matters and how reporting can be misread
Several pieces of coverage and commentary blur the line between “ICE’s detainees are largely non‑criminal” and “ICE employees are criminals,” which are fundamentally different claims; the available documents repeatedly focus on detainee composition and policy intent (e.g., “worst of the worst” rhetoric) rather than internal personnel records or misconduct prosecutions [3] [8] [5]. Some reporting about misconduct by federal agents describes controversies and local prosecutions of individual officers, but those accounts are case‑level and do not translate into a workforce‑wide percentage of agents with criminal records [9] [10]. Where data are aggregated—TRAC, Cato, Stateline, MinnPost—they pertain to who ICE arrests and detains, not who it employs [1] [3] [6] [2].
4. Conclusion, limits of the record, and next steps for verification
There is no basis in these sources to state a percentage of ICE agents who have criminal backgrounds; the documents instead supply multiple, consistent estimates showing most people ICE detains in 2025–2026 had no recorded criminal convictions (figures range around roughly 67–74% no convictions in various datasets and windows, with about 5% violent convictions noted by Cato for a broader timeframe) [1] [3] [2]. To answer the original personnel question would require access to ICE human‑resources or law‑enforcement personnel records, Office of Inspector General summaries, background‑check statistics, or FOIA disclosures—materials not present among these sources—so any percentage offered without those would be speculative [5] [9].