What the rule actually says, how the rulemaking will unfold, why it will be sued, and what employers should do now

On August 25, 2026, the Department of Homeland Security is publishing a notice of proposed rulemaking that would add a $103,265 fee to every H-1B cap-subject petition, payable at the time of filing, on top of every other fee an employer already pays. The proposal appeared on public inspection on August 24, 2026, under RIN 1615-AD20, CIS No. 2861-26, DHS Docket No. USCIS-2026-0298.

This is not an adjustment to the existing H-1B filing fee. It is a new, standalone charge that would be codified at 8 CFR 106.2(a)(3)(xii), and it is more than one hundred times the current base filing fee for an H-1B petition. If finalized as written, it would be the single largest fee ever imposed on an immigration benefit request in the history of the agency.

Below is a breakdown of what the rule does, how the rulemaking process will play out, the litigation that is almost certainly coming, and the practical impact on employers and workers.

I. What the Rule Proposes

The fee itself

DHS would charge $103,265 per H-1B cap-subject petition. The fee applies to petitions filed under the regular 65,000 cap and to those filed under the 20,000 advanced degree exemption at INA 214(g)(5)(C). It does not apply to cap-exempt petitions, extensions, amendments, transfers for beneficiaries already counted against the cap, or any other I-129 classification.

The exemption for cap-exempt filings is significant. Universities, related nonprofit entities, nonprofit research organizations, and governmental research organizations would be untouched. DHS explains this choice by analogy to the treatment of nonprofits and educational institutions under the Asylum Program Fee.

The fee is in addition to everything else: the base I-129 fee, the ACWIA fee, the fraud prevention and detection fee, the Asylum Program Fee, the H.R.1 fees codified in the April 2026 interim final rule, premium processing if elected, and any payment required under Presidential Proclamation 10973 if that payment is ever reinstated. DHS states expressly that an employer subject to both would pay both.

How DHS arrived at the number

The arithmetic is disclosed in a footnote and is unusually blunt. DHS identified total costs to be recovered of $8,777,488,035, divided that figure by an assumed annual volume of 85,000 cap-subject petitions, and arrived at $103,264.57, which it rounded to the nearest $5 increment.

In other words, the fee is not a measure of what it costs USCIS to adjudicate an H-1B petition. It is a revenue target divided by a headcount.

Where the money goes

This is the part of the rule that departs most sharply from prior practice. Historically, USCIS fee rules recovered USCIS costs. This rule would use H-1B cap fees to fund six agencies across three cabinet departments:

Agency or Department	Share	Allocation
USCIS	34.2%	$3,000.0 million
EOIR (DOJ immigration courts)	33.7%	$2,956.9 million
DOL	13.8%	$1,210.4 million
ICE	11.9%	$1,050.0 million
DOS	5.5%	$484.0 million
CBP	0.9%	$76.2 million

 

Some of the specific line items deserve attention, because they will be the center of gravity in any lawsuit:

• EOIR would receive nearly $3 billion, including funding for 8,400 new positions built around immigration judge teams, plus court space, guard costs, interpretation, transcription, and FOIA support. These are removal proceedings, not benefit adjudications.
• ICE would receive $900 million for “vetting of aliens pending adjudication,” $100 million for “vetting of applicants for admission,” and $50 million to cover a projected SEVP shortfall in excess of SEVIS fee revenue.
• DOS would receive funds for consular vetting systems, fraud programs at 270 consular posts, and also for refugee program costs, including Resettlement Support Center medical exams, refugee travel to the United States, International Organization for Migration operations, and expenses tied to Mission South Africa.
• DOL would receive $350 million for the Wage and Hour Division and $350 million for the Office of the Solicitor, plus program funding for PERM, prevailing wage, H-2A, and H-2B/CW-1.

An H-1B employer in Houston would, under this rule, be funding H-2A labor certification processing, refugee airport transfers, and immigration court capacity for removal cases.

The stated rationale

DHS offers two justifications. The primary one is revenue: the agency says INA 286(m) and 286(n), 8 U.S.C. 1356(m) and (n), authorize it to set fees recovering the “full costs” of providing adjudication and naturalization services, and to reimburse “any appropriation” for those expenses, and that nothing in the text limits recovery to USCIS costs alone.

The second is ability to pay. DHS reasons that cap-subject H-1B petitioners are “most willing and able” to bear the cost, citing a median FY 2025 H-1B compensation of $133,000 and an “unadjusted median wage of $798,000” over a six-year admission period.

DHS also asserts an “indirect benefit,” namely that employers would be less likely to hire an H-1B worker over a qualified American worker unless the need is legitimate.

II. The Rulemaking Process

Comment period

This is a notice of proposed rulemaking under 5 U.S.C. 553, not an interim final rule, so nothing takes effect on publication. Comments are due 30 days after publication, which points to on or about September 24, 2026. Confirm the exact date against the published Federal Register version, since the public inspection copy carries a bracketed placeholder.

Comments must be filed through regulations.gov under DHS Docket No. USCIS-2026-0298. Emails and letters to DHS officials do not count as comments. USCIS is not accepting mailed comments or digital media.

A 30-day window on an economically significant rule with an $8.8 billion annual price tag is aggressive. Executive Order 12866 contemplates 60 days as the norm for significant rules. That compression is itself a live procedural issue, and it is worth preserving in a comment.

What effective comments look like

Judicial review of a fee rule is record-bound. Arguments not raised in comments are frequently treated as forfeited. Employers and trade groups who want the strongest posture in later litigation should be submitting comments with:

• Company-specific data on how many cap petitions they file annually and what the fee would do to that number
• Evidence contradicting the 85,000 volume assumption
• Evidence on small entity impact, since the IRFA is thin
• Documentation of positions that would go unfilled, be offshored, or be eliminated
• Any evidence that the fee bears no relationship to the cost of adjudicating an H-1B petition

Anonymous or form comments carry little weight. Detailed, data-backed comments from actual petitioners carry a great deal.

What comes after

DHS must review comments and respond to significant ones in the preamble of a final rule. Given the scale, expect tens of thousands of comments. Realistic timeline for a final rule is several months at minimum, with OIRA review under E.O. 12866 required again at the final rule stage.

The timing question that matters most to employers is whether a final rule becomes effective before the FY 2028 cap registration period, which would ordinarily open in early March 2027. That is a tight but achievable window if DHS moves quickly, and there is every indication it intends to.

Note also that the fee attaches at the petition stage, not the registration stage. Registration would remain comparatively cheap. That structure creates an obvious dynamic: employers register, get selected, then decline to file. USCIS would then need to run additional selections to reach the numerical allocations, and the 85,000 revenue assumption starts to look fragile.

III. Probable Litigation

If this rule is finalized in anything close to its current form, it will be challenged. The only real questions are who files first and where.

Theory one: this is a tax, not a fee

This is the strongest challenge and it is already half-litigated. Under National Cable Television Association v. United States, 415 U.S. 336 (1974), and FPC v. New England Power Co., 415 U.S. 345 (1974), a charge qualifies as a fee only when it confers a specific benefit on the payer. A charge that funds general public benefits is a tax, and taxing authority must be delegated by Congress with a clear statement.

Apply that framework here. What specific benefit does an H-1B cap-subject petitioner receive from 8,400 new immigration judge positions adjudicating removal cases? From refugee medical exams? From H-2A labor certification processing? From IOM airport transfers? DHS’s own tables answer the question: roughly 66 percent of the revenue goes to agencies and programs with no adjudicatory connection to the payer.

DHS’s response is that INA 286(m) is a specific statutory authorization that displaces the general user-fee framework, and that 286(n) permits reimbursement of “any appropriation.” That reading has never been tested at this scale, and it proves too much. Under DHS’s theory, there is no ceiling: the agency could set an H-1B fee at $500,000 to fund the entire federal immigration apparatus.

Theory two: statutory authority under 286(m)

Even accepting that 286(m) governs, its text authorizes fees for “providing adjudication and naturalization services.” Removal proceedings before an immigration judge are not adjudication of a benefit request. Worksite enforcement is not an adjudication service. Refugee resettlement travel is not an adjudication service. Plaintiffs will argue DHS has read the operative limitation out of the statute.

There is a strong structural argument as well. Congress has repeatedly and specifically legislated H-1B fee amounts: the ACWIA fee, the fraud prevention and detection fee, the 9-11 biometric fee, and most recently the fees in H.R.1. When Congress sets specific dollar figures for a program again and again, the natural inference is that it did not silently authorize the agency to impose a $103,265 charge on the same program through a general cost-recovery provision.

Theory three: appropriations and fiscal law

Routing roughly $5.7 billion in fee revenue to DOJ, DOS, and DOL raises Appropriations Clause and Miscellaneous Receipts Act concerns, 31 U.S.C. 3302, along with longstanding anti-augmentation principles. DHS says it will execute interagency reimbursement agreements, but the details “would be determined before a final rule takes effect.” An agency asking a court to bless a multi-billion dollar interagency transfer whose mechanics have not been disclosed in the proposal is asking for trouble.

Theory four: major questions

An $8.8 billion annual reallocation, effectively restructuring how a substantial share of the federal immigration system is funded, based on a general cost-recovery provision enacted in 1988 and amended in 1990, is a textbook major questions argument under West Virginia v. EPA.

Theory five: arbitrary and capricious

This is where the rule is most vulnerable, and there are several independent problems.

The internal contradiction. DHS assumes 85,000 petitions will be filed and paid. It simultaneously claims the fee’s benefit is that employers “would be less likely to hire an H-1B worker,” and concedes registrations and filings may fall. These cannot both be true. If the deterrent works, the revenue does not materialize and the cost recovery fails. If the revenue materializes, the deterrent did not work and the stated benefit is illusory. Courts have set aside rules for less. Notably, this is also well beyond the statutory authority of USCIS.

The volume assumption is unsupported. FY 2025 cap-subject receipts were 106,711, and the two-year average was 108,910, because USCIS over-selects to account for abandoned selections. Assuming a clean 85,000 filings at $103,265 each, with no attrition modeling, is not analysis.

No benefits. The A-4 accounting statement lists monetized benefits as not estimated. E.O. 12866 and E.O. 13563 direct agencies to select approaches that maximize net benefits. A rule with $74.9 billion in ten-year costs and an empty benefits column invites the question of how the agency determined the benefits justify the costs.

Ability to pay is doing too much work. DHS’s median-wage math treats a $133,000 salary as evidence of $798,000 in employer capacity over six years. That conflates the employer’s payroll obligation with discretionary funds available for a filing fee. It also ignores that a small employer filing three cap petitions faces a $309,795 outlay before a single worker starts. The Borjas paper estimates a wage gap; DHS converts that into a willingness-to-pay finding for every petitioner in the country, from a fifteen-person startup to a Fortune 100 company.

Alternatives. DHS says it considered spreading costs across all benefit requests and rejected that as burdensome on lower-resourced applicants. It does not appear to have meaningfully considered intermediate options: a tiered fee, a phase-in, a small-employer reduction of the sort it built into the Asylum Program Fee, or a fee calibrated to actual adjudication costs.

Theory six: pretext

DHS concedes the fee is “close to” the $100,000 payment under Proclamation 10973, a policy that Judge Sorokin vacated on June 8, 2026 as an unconstitutional tax and an APA violation, and that the First Circuit declined to stay on July 24, 2026. The government’s merits appeal remains pending, with briefing expected to conclude in October 2026.

Plaintiffs will argue that this rulemaking is an attempt to accomplish through 286(m) what the proclamation could not accomplish through 212(f), and that the stated cost-recovery rationale is a post hoc justification for a predetermined number. Department of Commerce v. New York supplies the framework for a court asked to look behind a stated rationale that does not match the record.

Timing and posture

A challenge cannot be filed until a final rule issues; the NPRM is not final agency action. Expect complaints filed within days of publication of a final rule, motions for preliminary injunction or expedited summary judgment, and a request for vacatur under 5 U.S.C. 706. Given the cap registration calendar, plaintiffs will press hard for relief before the FY 2028 filing window.

Employers should also watch the refund question. If the fee is collected and the rule is later vacated, the mechanics of getting $103,265 back per petition are not addressed in the proposal in any way that should give employers comfort. The unresolved refund mess from the proclamation payments is instructive.

IV. What Employers Should Do Now

1. Do not panic and do not assume this is law. Nothing has taken effect. A proposed rule is a proposal.
2. Comment, with data, before the deadline. Docket USCIS-2026-0298 on regulations.gov. Specific, documented, company-level comments are what move the needle and what preserve arguments for later.
3. Model FY 2028 scenarios now. Build budgets for both outcomes. Identify which roles you would still sponsor at $103,265 and which you would not.
4. Audit your workforce for alternatives. Identify who might qualify for O-1A, NIW, TN, E-3, H-1B1, or other options. Start those assessments early, not in March.
5. Watch California v. Mullin. The First Circuit’s merits decision, expected after briefing concludes in October 2026, will shape the legal terrain this rule lands on.
6. Preserve records. If you abandon a cap selection, decline to sponsor, or lose a candidate because of anticipated fees, document it. That evidence matters for comments and for standing later.

This article is current as of August 24, 2026, and reflects the public inspection version of the proposed rule. It is provided for general informational purposes and is not legal advice. Fee amounts, deadlines, and litigation posture are subject to change.

By: Steven Brown

Steven A. Brown is a Partner at Reddy Neumann Brown PC, where he leads the firm’s Litigation Team, addressing delays and denials of immigration benefits, FOIA requests, and policy and regulatory challenges. Steven is dedicated to delivering practical and effective solutions for clients facing unreasonably delayed or unlawfully withheld immigration benefits, including Employment Authorization Documents (EADs), advance parole, green cards, 221(g) decisions, EB-5 delays, and other immigration-related matters. His litigation efforts were instrumental in Shergill, et al. v. Mayorkas, a landmark case that led to the U.S. government recognizing that under the INA, L-2 and E visa spouses are authorized to work incident to their status, eliminating the need for separate EAD applications. This case has transformed work authorization for thousands of families across the United States.