Do Big H-1B Sponsors Announce Big Layoffs? A WARN Reality Check

There is a common assumption that the companies filing the most H-1B paperwork are also the ones cutting the most jobs. It sounds intuitive: the biggest tech and consulting names dominate the headlines both for visa sponsorship and for layoffs. But when you actually line up the two datasets, the overlap is far smaller — and far messier — than the narrative suggests.

This is Part 3 of our Layoffs × H-1B series. In the earlier parts we looked at how layoff announcements are tracked and how the timing lines up with visa filings. Here we do something more direct: we take the top 100 H-1B sponsors and check each one against WARN notices — the official layoff filings employers must submit to state labor agencies — in the states we track.

The short version: most big sponsors have no WARN notice at all, a handful have small ones, and a single company distorts the entire picture.

What the numbers actually show

As of our August 2026 snapshot, only 22 of the top 100 H-1B sponsors have any WARN notice on record in the tracked states (California, Texas, New Jersey, Illinois, Massachusetts, and Virginia). The other 78 — including many of the largest filers — simply do not appear.

Of the 22 that do, most notices are small relative to how many Labor Condition Applications (LCAs) the company files. A firm can submit tens of thousands of LCAs over its lifetime and show up in the WARN data with a notice covering a few hundred workers. The two datasets are measuring different things at different scales, and the mismatch matters.

The chart above shows the sheer filing volume of the largest sponsors. Keep that scale in mind as we walk through the layoff side — because almost none of the WARN notices come close to it.

The one giant exception: Tesla

Every rule has an outlier, and here it is Tesla. A single WARN notice filed in Austin, Texas in April 2024 named roughly 2,688 workers — the largest tech layoff notice in this entire dataset. For context, Tesla has about 10,903 lifetime LCA filings, so this one notice is genuinely large in both absolute and relative terms.

Tesla is the exception that proves the pattern. When people picture "big H-1B sponsor announces big layoff," they are essentially picturing this one event. It is real, it is documented, and it is unusual. No other top sponsor in the tracked states has a single notice on anything like that scale.

Everyone else is comparatively small

Once you move past Tesla, the WARN notices shrink dramatically. Here is what the rest of the field looks like among big sponsors:

• Meta: 5 notices covering roughly 336 workers (November 2022 to May 2023).
• Accenture: 3 notices, about 971 workers (2023–2025).
• Tata Consultancy Services: 3 notices, about 200 workers.
• Intel: 3 notices, about 416 workers (most recent July 2025).
• T-Mobile: 7 notices, about 399 workers.
• JPMorgan Chase: 7 notices, about 633 workers — spanning 2007 all the way to June 2026, the longest arc in the set.

JPMorgan's record is the most interesting shape here. Rather than one dramatic event, it is a nearly two-decade drip of routine, mostly modest notices. That is what a large employer's WARN history usually looks like: periodic restructurings, office closures, and reorganizations, not a single catastrophic cut. It is also a reminder that a WARN notice is a normal part of operating at scale, not necessarily a sign of distress.

Fresh 2026 notices worth watching

The dataset is a living thing, and a few big sponsors picked up new notices in 2026:

• Samsung Electronics America: roughly 739 workers (New Jersey, July 2026).
• Salesforce: about 74 workers (August 2026).
• Citibank: 2 notices, about 127 workers (June–August 2026).
• Uber: about 41 workers (July 2026).

These are recent enough that they may still be evolving. But notice the scale again — outside of Samsung's mid-sized New Jersey notice, these are small numbers against companies that file thousands of LCAs.

The blind spot you have to know about

Here is the part that should reshape how you read everything above: Washington State is not in our tracked set.

That is not a minor caveat. Two of the most heavily reported layoff stories in tech — cuts at Microsoft and Amazon, both headquartered in Washington — are essentially invisible in this dataset. If you searched these WARN records for the companies whose layoffs dominated the news, you might wrongly conclude they had cut almost no one. They didn't; we simply aren't looking in the state where those notices land.

The lesson is blunt: absence from this dataset is not evidence of no layoffs. WARN coverage is geographically incomplete here, and even where it is complete, WARN law only applies to larger employers above specific event-size and notice thresholds. Small rounds of cuts, cuts spread thinly across many sites, and cuts in untracked states can all slip through entirely.

Why the two datasets don't line up

It is tempting to draw a straight line from "files a lot of H-1Bs" to "lays off a lot of workers," but the data resists it for several structural reasons.

First, an LCA is not a person and not a visa. A Labor Condition Application is a wage-and-conditions filing an employer submits before it can sponsor or extend an H-1B worker. A company can file many LCAs per worker over time, and a certified LCA never guarantees a hire or a USCIS approval. So filing volume overstates headcount.

Second, WARN headcounts are company-wide. When a notice says 2,688 or 336 workers, those are total affected employees at that site or event — not H-1B workers specifically. Nothing in the WARN data tells you the visa status of anyone laid off. We make no claim, and you should make none, that any particular H-1B worker was affected.

Third, the timing and geography rarely align. LCA filings are national and continuous; WARN notices are state-specific and event-driven. A sponsor's filings might concentrate in California while its layoffs, if any, register in a state we don't track.

How to read this honestly

If you are a worker weighing an offer, a researcher, or a journalist, the practical takeaways are narrow but useful:

• Big sponsor ≠ big layoffs. Only about a fifth of the top 100 sponsors show any WARN notice in the tracked states, and most are small.
• Tesla is a genuine outlier, not a template. Do not generalize from it.
• A long WARN history (like JPMorgan's) is normal for large, decades-old employers and doesn't imply instability.
• Trust the news over silence. If the press reports major cuts at a Washington-based company, believe the reporting — this dataset just can't see it.

None of this establishes causation between visa sponsorship and layoffs, and it isn't meant to. It is a reality check against a tidy story that the underlying records don't actually support. The overlap between the biggest H-1B filers and the biggest documented layoffs is thin, uneven, and — outside of one Austin notice — surprisingly quiet.

A note on the data: FY2026 H-1B figures are partial (roughly two quarters reported). LCA filings are not visa approvals, a certification rate is not a USCIS approval rate, and a withdrawn application is not a denial. WARN headcounts are company-wide estimates and are never H-1B-specific.