Worst of the worst? Most US immigrants targeted for deportation in 2025 had no criminal charges, documents reveal

The Guardian - World NewsEN 5 min read 100% complete by Will CraftFebruary 22, 2026 at 02:00 PM

AI Summary

long article 5 min

A Guardian analysis of government documents revealed that in 2025, the majority (77%) of individuals facing initial deportation proceedings had no criminal convictions. This finding contradicts the Trump administration's claim that its immigration enforcement focused on deporting serious criminals. The analysis, based on nearly 140,000 I-213 forms from January to mid-August 2025, showed that only 23% of those targeted for deportation had a criminal conviction. Of those with convictions, nearly half were for non-violent traffic or immigration offenses. The documents, obtained through a Freedom of Information Act lawsuit, offer a detailed look at the criminal records of those targeted for deportation during that period.

Key Entities & Roles

Key Player Opposition Context

Reporters Committee for Freedom of the Press 40% The Guardian ×350% Immigration and Customs Enforcement (ICE) 60% United States 70% Department of Homeland Security (DHS) ×680% Donald Trump ×290%

Keywords

deportation 100% criminal conviction 80% immigration enforcement 80% i-213 forms 70% dhs 70% traffic offenses 60% non-violent offenses 60% arrests 50% immigration offenses 50% guardian analysis 40%

Sentiment Analysis

Very Negative

Score: -0.60

Source Transparency

Source

The Guardian - World News

Classification Confidence

90%

This article was automatically classified using rule-based analysis.

Topic Connections

Explore how the topics in this article connect to other news stories

Network visualization showing 4 related topics

View Full Graph

deportation criminal conviction immigration enforcement i-213 forms dhs traffic offenses non-violent offenses arrests immigration offenses guardian analysis

Explore Full Topic Graph

Find Similar Articles

AI-Powered

Discover articles with similar content using semantic similarity analysis.