Why ICE Exists
On January 28, 2026, the United States Department of Homeland Security released a short video honoring the life of Megan Bos, an Illinois woman whose death would otherwise have remained one more private tragedy in an age drowning in them. The timing was deliberate. The release came one year into the Trump administration, at a moment when immigration enforcement had ceased to be an embarrassed afterthought and was once again treated as a foundational obligation of the state. The message was not emotional. It was forensic. Megan Bos did not die because America lacks laws. She died because the state failed to enforce them when it mattered.
One year into the Trump administration, the January 28 DHS report should be read not as a memorial, but as a verdict.
Bos’s case involved no statutory uncertainty. The accused had been arrested, processed, and placed inside the criminal justice system. He was later released under Illinois’ SAFE-T Act, which sharply limits pretrial detention and reflects a policy judgment that custody itself should be exceptional. The release was not the result of a finding that the defendant posed no danger. It was the product of a framework that elevates release as a default. Federal authorities intervened later. By then, delay had already imposed its cost.
The case illustrates a broader point often obscured in contemporary debate: non-enforcement is not a neutral condition. It is a decision. When an individual who is removable under federal law is released rather than transferred into federal custody, risk is not eliminated. It is reassigned — from institutions designed to manage it to the public, which never agreed to assume it. Megan Bos bore the consequence of that reassignment.
This is the institutional role of U.S. Immigration and Customs Enforcement. ICE is not a substitute for local policing or courts. It is the only authority with the mandate to remove non-citizens once state and local processes conclude. In a fragmented system — where police arrest, courts adjudicate, and legislatures set policy-ICE functions as the final integrating mechanism. When that mechanism is resisted or disabled, risk does not disappear. It circulates.
Recent events in Minnesota offer a clear illustration of how this dynamic plays out. Over the winter, federal agents expanded targeted enforcement operations in Minneapolis and the surrounding Twin Cities region, prompting protests, political backlash, and public clashes between local officials and federal authorities. The controversy focused largely on optics and tactics. Far less attention was paid to the profiles of those actually being arrested.
According to disclosures by DHS and contemporaneous reporting, individuals taken into federal custody during these operations included convicted perpetrators of aggravated sexual crimes involving minors, individuals with criminal sexual conduct convictions, offenders convicted of drug trafficking, and others with documented histories of domestic violence and assault. In one case, ICE arrested an individual subject to a federal warrant for the sale and distribution of amphetamines, alongside prior convictions that had not prevented his continued residence in the community.
These were not first-time offenders, nor individuals awaiting adjudication. They were people who had already passed through local systems and been released — sometimes repeatedly — despite criminal histories already established. The Minnesota arrests did not create new risks. They revealed accumulated ones.
This pattern is not confined to one state. During the Biden administration, independent and federal estimates converged on a population of roughly 10 to 11 million people living in the United States without legal status, many released pending proceedings, many never located again, and many existing outside any continuous monitoring framework.
At that scale, assurances that “most are law-abiding” lose analytical relevance. Risk in large systems does not operate on percentages. It operates on absolute numbers. A system that loses track of millions is not exercising compassion; it is forfeiting visibility. No serious regulator would tolerate comparable opacity in aviation safety, financial markets, or public health. Yet in immigration enforcement, diminished situational awareness has often been defended as a moral achievement.
ICE exists because scale changes responsibility. Local courts lack authority to remove non-citizens. States lack jurisdiction. Cities lack both capacity and mandate. Interior enforcement at this magnitude is not an ideological preference. It is a functional requirement of governance. When that requirement is treated as optional, the consequences follow predictably.
Where enforcement is obstructed, failure follows. In December 2025, DHS publicly rebuked Fairfax County, Virginia, after local officials declined to honor an ICE detainer request and released a criminal illegal alien who committed murder the following day. DHS stated plainly that the killing was preventable.
Similar failures have been documented elsewhere. In Minnesota and beyond, ICE arrests have included convicted child rapists and killers who had been living freely because local jurisdictions declined to cooperate with federal custody transfers. These were not clerical oversights. They were policy choices, repeated, defended, and predictable in outcome.
Nationwide enforcement operations during the same period identified what DHS described as the “worst of the worst”: murderers, violent gang members, and child sex offenders apprehended only after slipping through local enforcement gaps. In one case, an Egyptian national accused of killing an elderly woman with an improvised incendiary device at a public gathering was detained only after cooperation had already collapsed.
Even where ideology recedes, reality intrudes. In late 2025, federal prosecutors detailed the case of a twice-deported illegal immigrant who unlawfully re-entered the United States and later caused a fatal car crash that killed two teenagers. Removal orders had been issued. The law was settled. Enforcement was not sustained.
The numbers tell the same story. In 2025, ICE conducted nearly 600,000 administrative arrests and carried out more than 600,000 removals, the highest one-year totals on record. According to DHS, roughly 70 percent of those arrested had prior criminal convictions or pending charges. This was not indiscriminate enforcement. It was triage.
Where enforcement was obstructed, risk accumulated. Between 2022 and early 2025, more than 25,000 ICE detainers were declined nationwide. California accounted for more than half. Boston acknowledged ignoring every detainer request it received in 2025. These refusals did not eliminate risk. They displaced it.
The final illusion is that enforcement is punitive rather than preventative. Yet as removals increased, DHS also recorded nearly two million voluntary departures — individuals choosing to leave rather than test a system that had resumed functioning. Deterrence rarely announces itself. It is measured in crimes that do not occur, victims who never come into being, funerals that are never held.
Megan Bos’s death followed a familiar sequence: arrest, release, delay, irreversible harm. Every step complied with prevailing policy norms. Together, they proved fatal.
ICE intervened too late not because it was unnecessary, but because it had been prevented from acting sooner.
The January 28 DHS report should therefore be read not merely as a memorial, but as a verdict. A state that declines to enforce its own laws does not become more humane. It becomes less responsible. ICE exists because modern societies cannot govern through moral outsourcing alone. When enforcement is treated as discretionary, the costs do not vanish. They accumulate-until someone else pays them.
Megan Bos was not an anomaly.
She was the invoice.
READ MORE from Kevin Cohen:
Drug Gangs, Child Gunmen and Antisemitic Abuse — Welcome to Marseille