DHS "Worst of the Worst" Website: Rife with Errors

On February 19, 2026, CNN published an analysis revealing that hundreds of people listed on the DHS website were incorrectly characterized. Many were listed for single charges of traffic offenses, marijuana possession, or illegal reentry—not the violent crimes the "Worst of the Worst" branding implied.^[1][2] After receiving CNN's questions, a DHS spokesperson admitted that "the charges against hundreds of immigrants listed on the website were described incorrectly by the agency."^[1]

DHS attributed the errors to a "glitch" affecting "about 5% of the entries," claiming many listed for traffic offenses actually had "additional crimes."^[3] However, 5% of 25,000 entries equals approximately 1,250 people falsely characterized—and DHS provided no technical explanation for how a "glitch" could systematically misrepresent charges. The agency claimed the "glitch" was "resolved" within one day of CNN's inquiry, an implausibly short timeline for complex database systems requiring data validation.^[3]

Former acting ICE director John Sandweg noted that "the vast majority of so-called criminal aliens are individuals charged with or convicted of traffic offenses, DUIs and immigration-related offenses"—context conspicuously absent from DHS promotional materials.^[1] Jason Houser, former ICE chief of staff during the Biden administration, stated bluntly: "It just shows that this is about the political debate. It's not about actually arresting the most convicted criminals."^[15]

CNN's analysis revealed another deceptive pattern: cities with the highest arrest numbers were small towns with large federal detention facilities.^[13] The city representing the most arrests on the DHS website was Conroe, Texas—population approximately 114,000—which is home to the Joe Corley Processing Center, a privately owned detention facility that ICE uses to house immigrants. Similar patterns appeared in Eden, Texas; Lompoc, California; and Yazoo City, Mississippi—all small towns with federal detention centers.

This indicates DHS counted routine transfers from federal prison custody as new "arrests" of community threats, inflating the narrative that dangerous criminals were "lurking in your neighborhood." The geographic manipulation allowed DHS to claim high arrest numbers in specific communities while concealing that these "arrests" were actually administrative transfers of people already in federal custody—sometimes for years.

Minnesota Corrections Commissioner Paul Schnell accused DHS of "pure propaganda" and presented video evidence refuting DHS claims about Operation Metro Surge arrests.^[6][7] While DHS claimed it "arrested" people released into communities by Minnesota Department of Corrections, Schnell showed videos of coordinated, routine transfers to ICE custody—not "arrests."

Some people on DHS lists were transferred to ICE years ago—as far back as the 1990s. Many listed were never in Minnesota prisons or had no criminal record in Minnesota.^[6][28] Governor Tim Walz similarly accused DHS of inflating numbers by taking credit for pre-existing state custody transfers that occurred before Operation Metro Surge even began.^[29]

Schnell stated: "This is no longer simple misunderstanding. At best, DHS fundamentally misunderstands Minnesota's correctional system... At worst it is pure propaganda, numbers released without evidence to stoke fear rather than inform the public."^[6]

Multiple independent analyses directly contradict DHS's public messaging that it was targeting "the worst of the worst." TRAC Immigration's detention tracker has repeatedly shown that roughly seven in ten people held in ICE detention have no criminal conviction at all—most recently 70.6% (46,436 of 65,765 detainees).^[4] Separate reporting from around the time the "Worst of the Worst" website was being promoted found the same pattern, undercutting DHS's branding of the people it was showcasing.

The Cato Institute, analyzing leaked DHS data from October 1, 2025 to the present, found that 73% of people booked into ICE custody had no criminal conviction, only 8% had violent or property crime convictions, and only 5% had violent criminal convictions.^[5] When the Cato Institute published this analysis, DHS called the report's main chart "made up" with "no legitimate data behind it"—yet the raw data came from DHS itself.^[16]

Between September 21, 2025 and January 7, 2026, ICE detention increased by 11,296 people. Of that increase, only 902 were convicted criminals (8%), 2,273 had pending criminal charges (20%), and 8,121 had no criminal convictions or pending charges (72%).^[12] This means 92% of ICE detention growth came from people with no criminal convictions—the exact opposite of the "worst of the worst" narrative.^[12][24]

The website errors are part of a documented pattern of DHS making false claims, then retracting or being contradicted by evidence. On January 24, 2026, Alex Pretti—a U.S. citizen—was shot and killed by federal agents in Minneapolis. Officials immediately called him a "domestic terrorist" who "brandished" a gun and "wanted to massacre law enforcement."^[8][9][27]

Multiple bystander videos showed Pretti's hands holding only his mobile phone when shots were fired—no brandishing of a weapon. Footage captured an officer removing a handgun from Pretti's waist area, then moving away—after which the first shot was fired.^[8] Ten shots total were fired, including rounds after Pretti was on the ground motionless.^[9] Video analysis by the open-source investigations outlet Bellingcat found the slide of Pretti's own firearm did not appear to move to the rear in the footage—consistent with it not having been fired—though Bellingcat and other reviewers cautioned no forensic ballistics report had been publicly released to confirm this.^[25]

Former DHS officials warned this pattern erodes public trust. John Sandweg, former acting ICE director during the Obama administration: "Public trust is everything to these agencies—once destroyed, it affects all their work."^[15]

The website errors exist within a broader context of ICE data integrity problems documented by the U.S. Government Accountability Office. In July 2024, the GAO found that ICE understates the total number of people it detains in its public reports.^[17] ICE excludes people who were first booked into certain temporary facilities before being transferred to immigration detention facilities—an exclusion that amounts to tens of thousands of people.

During calendar year 2022, ICE did not count among its total detained population 203,350 individuals—roughly 42% of people detained—because they initially were held in temporary ICE or CBP facilities. ICE claimed it did not count these individuals because it did not detain them for very long. However, the GAO investigation found that approximately 70% of individuals not counted in the tally were eventually detained by ICE for weeks, and in some instances, even years.^[18]

The GAO recommended ICE publicly report complete detention data and explain methodology. DHS did not concur, stating ICE already reports sufficient information. The GAO continues to believe ICE should report complete and transparent information.^[17] A subsequent 2025 GAO review found DHS inspection programs for ICE detention facilities lack clear performance goals and measures, and that inspectors found 31 of 33 reviewed facilities failed to meet applicable standards.^[19]

In February 2026, a federal court granted emergency relief to 13 members of Congress, ordering DHS and ICE to restore unannounced congressional oversight of ICE detention facilities.^[20] After a court had preliminarily stayed a DHS policy requiring members of Congress to provide prior notice of oversight visits to ICE facilities in December, DHS Secretary Kristi Noem secretly reinstated the requirement through a previously undisclosed memorandum.

On February 3, 2026, Representative Robert Garcia and Senator Richard Blumenthal held a bicameral public forum to receive testimony on the violent tactics and disproportionate use of force by agents of the Department of Homeland Security.^[21] Senate Democrats demanded Secretary Noem testify "without delay" following fatal shootings by federal immigration officials in Minneapolis. Noem has failed to appear before the Senate Homeland Security Committee for its annual hearing since her confirmation, despite testifying before a House panel.^[26]

The Washington Post reported that DHS is targeting Americans with administrative subpoenas—unlike judicial subpoenas approved by a judge, administrative subpoenas are issued by federal agencies, allowing investigators to seek information about individuals from tech and phone companies without a judge's oversight.^[22]

The DHS website errors represent more than a technical failure—they exemplify a broader pattern of government misinformation used to justify aggressive enforcement policies. Hundreds of people were publicly characterized as "worst of the worst" criminals when they had only minor charges, with their names, photos, and locations searchable online. This false characterization had real human consequences.

The website was used to justify Operation Metro Surge deploying 3,000 federal agents to Minnesota,^[14] warrantless home entry policies, refugee detention policies, and other controversial enforcement actions. Policy justification based on false data undermines democratic accountability and informed public debate.

Former DHS officials warn that repeated false statements—about Alex Pretti, Renee Good, Marimar Martinez, and the website errors—undermine public trust in law enforcement. John Sandweg: "Public trust is everything to these agencies—once destroyed, it affects all their work."^[15]

The pattern of errors suggests the problem may be rooted in how DHS defines "criminal aliens" rather than a one-time technical glitch. DHS inflates numbers by counting pending charges (not convictions) as "criminals" and claiming this "doesn't even include foreign fugitives, terrorists, and gang members who lack a rap sheet in the U.S."^[16]

Despite GAO recommendations, court orders, and congressional pressure, DHS has resisted transparency measures and independent oversight. The website remains online with continued daily updates, but DHS has not published a corrected list of affected entries, technical explanation of what caused the errors, independent audit of remaining data accuracy, or methodology for validating criminal charge descriptions.

The Department of Homeland Security's admission that its "Worst of the Worst" website contained hundreds of errors falsely characterizing people with minor offenses as dangerous criminals represents a significant breach of public trust. DHS only acknowledged the errors after media investigation, offered an implausible "glitch" explanation with no technical details, and claimed it was resolved in one day.

Independent data contradicts the core premise of the website: the large majority of ICE detainees have no criminal conviction, only a small share have violent convictions, and the bulk of detention growth since September 2025 came from people with no criminal record. Minnesota officials provided video evidence refuting DHS claims about Operation Metro Surge arrests, showing routine custody transfers DHS later claimed as enforcement "arrests."

The website errors are part of a documented pattern of false official statements contradicted by video evidence in the Alex Pretti, Renee Good, and Marimar Martinez cases. Former DHS officials warn this pattern undermines public trust in law enforcement and affects all agency work. The immediate question for DHS is narrow: what exactly went wrong and how was the correction validated. For the public, the larger question remains structural: whether a database marketed as proof of precision targeting can earn credibility when the underlying data presentation proves fragile.