ICE Says It Arrested 1,328 People in Virginia and Maryland During a Two-Week Operation in Early August

DHS Says 1,328 Arrests, But Its Own Numbers Leave 930 Unexplained

The Department of Homeland Security announced on Aug. 24 that Immigration and Customs Enforcement arrested 1,328 people across Virginia and Maryland over two weeks in early August^[1]. DHS gave the operation a name built for the capital, "Operation Safe Community – Washington, D.C.," even though the arrests it described happened in the two surrounding states, not inside the District itself^[1]. The window ran from Aug. 1 to Aug. 14. DHS didn't say a word about it publicly until 10 days after the operation had already ended^[1].

Buried in that same release is the number both sides will spend the next few weeks arguing over. DHS said almost 400 of the 1,328 people arrested had been convicted of or charged with a crime in the United States^[1]. Simple subtraction does the rest: that leaves roughly 930 people, about 7 in 10 of the total, with no criminal conviction or charge attached to their arrest^[1]. DHS didn't highlight that number. It highlighted the other one, describing arrests for rape, kidnapping, robbery, drunk driving and identity theft, and pointing to people it says belong to MS-13, the 18th Street Gang or Tren de Aragua^[1].

Both of those figures came from the same three-paragraph press release, and neither has been checked by anyone outside DHS. No jurisdiction breakdown. No case list beyond a handful of named examples. No data file^[1]. That gap is where the fight over what this operation actually was is going to play out.

What "Convicted or Charged" Actually Bundles Together

It matters that DHS's headline figure isn't 400 convictions. It's 400 people convicted or charged, and those are two very different things. A conviction means a court found someone guilty or they pleaded guilty. A charge means someone accused them of a crime, and that accusation hasn't been tested in court. DHS folds both into one number, so even the "almost 400" isn't a clean measure of proven criminal history^[1].

That's why the roughly 930 people left over matter so much to critics. DHS's own position is that everyone it arrested was already breaking the law simply by being in the country without legal status, so a non-criminal arrest isn't a mistake — it's the statute doing what it's supposed to do^[1]. Immigrant-rights groups read the same ratio the opposite way. They argue the named violent cases are what gets publicized, while the bulk of the arrests are ordinary people swept up along the way^[4].

Independent data, as far as it goes, tilts toward the second reading. Researchers at the Deportation Data Project at UC Berkeley used Freedom of Information Act lawsuits to pry loose ICE's own arrest records for this region^[4][7]. Between Jan. 20, 2025, and March 10, 2026, ICE made about 19,500 arrests across D.C., Maryland and Virginia. Roughly 11,600 of those people, about 60%, had no prior criminal record^[4][7]. That's not this specific operation. It's the broader trend line this operation sits inside of.

The Mechanism Both Sides Actually Agree On

Underneath the argument over numbers is a piece of machinery that DHS and its critics both understand the same way, even though they blame each other for how it plays out. It's called a detainer. When ICE wants someone a local jail already has in custody, it can send that jail a detainer request, asking the jail to hold the person up to 48 hours past their release date so ICE can pick them up at the door. One officer, one handoff, done quietly.

Virginia and Maryland have each pulled back from that system over the past year. Gov. Abigail Spanberger rescinded a Youngkin-era order that required Virginia state police and corrections agencies to sign ICE cooperation deals, and ended the state prison system's transfer agreement with ICE^[10]. In May, she went further, signing an order that bars ICE from making warrantless arrests on Virginia state property^[19]. Gov. Wes Moore signed a Maryland law this year limiting local sheriffs from entering their own agreements to help ICE^[16].

Courts have generally held that states can decline to help enforce federal immigration law, even though federal law itself still applies inside their borders^[16]. What that means in practice is that declining a detainer doesn't stop an arrest from happening. It just moves the arrest from a jail hallway to a street, a home or a job site, done by ICE officers acting on their own^[1][14]. DHS says that shift explains why so many of its arrests don't come attached to a criminal record: agents doing a street arrest sweep up whoever else is present and removable, not just the one person they came for. Critics say that's exactly the harm they're trying to prevent — a system that's supposed to target specific threats instead catching bystanders.

There's a resource strain feeding this too. ICE's funding has grown much faster than its supply of trained officers. Nationally, daily arrests averaged 1,264 from December 2025 into January 2026, peaking near 1,456 a day in late January^[8]. Hitting numbers like that with a limited staff pushes agents toward the fastest available arrest, and the fastest arrest is whoever is standing there, not a hard-to-locate fugitive^[8][14].

Two Governors Who Haven't Said a Word Yet

Neither Spanberger nor Moore had publicly responded to this operation by the time the first news stories ran^[13]. That silence is doing a lot of work in how the story is landing. Their case, when they've made it before, rests on more than just politics. Under the Tenth Amendment, the federal government can't force state officers to carry out a federal program, and declining a detainer request is a choice courts have said states are allowed to make.

There's a public-safety argument underneath that legal one too. A detainer is a request, not a warrant signed by a judge, and a jail that holds someone past their release date on that request alone can be sued for it. Spanberger's team has also argued that if immigrant communities believe calling 911 might bring ICE to their door, victims stop calling, and crimes go unreported. Her May order tracks that logic directly, requiring a warrant before ICE can make an arrest on state property^[19].

This isn't the two governors' first run-in with DHS over this. In February, DHS put out a release timed to Spanberger's response to the president's State of the Union address, claiming she'd banned all Virginia law enforcement from working with ICE^[12]. A WRIC fact-check found that wasn't accurate — her order applied to state agencies, not every local department in Virginia^[11]. That history is part of why this new number lands as one more round in an ongoing fight rather than a standalone story.

Local sheriffs don't split neatly along the same lines as the governors and DHS. Frederick County's sheriff in Maryland warned that banning jail-based cooperation wouldn't reduce enforcement, just push it into neighborhoods as street arrests, which he considers more dangerous for his own deputies and everyone else^[16]. Other law enforcement officials argue the opposite: that participating in ICE programs burns local jail resources on federal work and costs departments the community trust they need to solve their own cases.

How the Story Got Told Differently on Each Side

Coverage of the same three-paragraph press release split hard along predictable lines. Fox News called it a "jaw dropping" number ICE "quietly" pulled off, leading with the worst offenses on DHS's list^[2]. Breitbart repeated DHS's crime list almost word for word and didn't mention the roughly 930 arrests without a listed conviction or charge, even though that number is simple subtraction from the same release^[6]. RedState's headline wasn't really about the arrests at all — it was about how angry the two governors were expected to be, even though the story itself notes neither had said anything yet^[13].

Local and national outlets closer to the center handled the number more cautiously. NBC Washington's headline attributed the figure directly to DHS rather than stating it as settled fact, and noted the remainder were immigration violators without a criminal record attached^[3]. WJLA, a Sinclair-owned Washington station, did the most independent work in the mix, using the FOIA-obtained ICE data to report that most regional arrests over the past year involved people with no criminal record^[4]. Even that framing was a choice: it foregrounded the non-criminal majority without giving equal space to DHS's argument that sanctuary policies are what pushed arrests onto the street in the first place.

Outside the U.S., the operation barely registered. No major non-U.S. outlet appears to have covered it as of Aug. 25. The closest external vantage point is a commentary from Real Instituto Elcano, a Madrid think tank, which reads the broader pattern of U.S. urban ICE raids as a story about an agency with more money than trained officers, adopting tactics that generate both arrests and backlash^[14].

What's Still Missing

DHS said this operation followed a similar one in Georgia in July, "Operation Safe Community – Atlanta," which it said produced more than 1,200 arrests^[1]. That suggests this approach, announce after the fact, lead with the worst cases, let the ratio go unexamined, is now a repeatable template rather than a one-off.

What isn't repeatable, at least not yet, is verification. Nobody outside DHS has seen a breakdown of who the 1,328 people are, where in Virginia or Maryland they were picked up, or how many of the gang affiliations DHS cited will hold up under scrutiny. Removal proceedings for people caught in this operation will likely take months to move through an already backlogged immigration court system. Until DHS releases more than a press release, or until the two governors decide to respond, the only numbers on the table are the ones the agency chose to publish.