How San Francisco Is Challenging Anti-DEI Federal Grant Terms

Risk posture as of July 27, 2026: San Francisco has filed or joined at least four challenges to anti-DEI conditions attached to federal grants. A practitioner synthesis describes the San Francisco line as having produced three preliminary injunctions and one temporary restraining order through mid-2026; the case-specific materials available for the four tracked matters below identify two named preliminary-injunction rulings, one TRO, and one DOE case still awaiting a ruling. That distinction matters. A TRO is not a merits judgment, a preliminary injunction is still provisional relief, and the undecided DOE case should not be treated as a judicial signal yet. [1]

The useful question is not whether “DEI” is lawful or unlawful in the abstract. For grant recipients, the live question is narrower and more operational: which agency condition was blocked, under what theory, and can that reasoning be carried into a different grant program before the next certification, drawdown, or termination deadline?

San Francisco anti-DEI grant-condition tracker

The table should be read conservatively. The dollar figures are exposure figures reported in litigation coverage, not judicial findings about entitlement or damages. The relief is also not uniform. A TRO preserves the status quo for a shorter emergency window; a preliminary injunction reflects a fuller early-stage showing but still does not decide the ultimate legality of the condition. The DOE case is filed but undecided, so it belongs in the risk map, not in the win column.

What the rulings actually have in common

The cleanest pattern is statutory and procedural before it is cultural. In the rulings that exist, courts have been receptive to the argument that agencies cannot add anti-DEI conditions to congressionally created grant programs unless Congress authorized that condition with enough clarity for recipients to know what they are accepting. O’Melveny’s practitioner synthesis frames the San Francisco cases around that same recurring statutory-authority problem: agencies attempted to condition federal financial assistance on DEI-related restrictions not found in the underlying grant authorizations. [1]

Judge Donato’s police-grant ruling is the most direct collision between agency condition and congressional text. The Community Policing Act does not merely leave hiring diversity unmentioned; the available coverage describes it as requiring diversity in hiring. That matters because DOJ’s anti-DEI pledge did not arrive as an implementation detail filling a statutory gap. In the San Francisco challenge, it was treated as running against the grain of the grant program Congress created. [2][3]

For a recipient-side lawyer, that is a stronger fact pattern than a generic objection to an administration’s priorities. If a grant statute instructs or contemplates one thing and an agency certification pushes recipients toward the opposite, the recipient can frame the dispute as one about delegated authority, not policy preference. The question becomes whether the executive branch is administering the grant Congress enacted or rewriting the price of participation after the money has been appropriated.

Judge Chen’s AmeriCorps TRO is different. The operative problem there was vagueness under the Spending Clause. SFUSD faced the possible loss of more than $650,000, described as about half the Healthy Choices Program budget, under conditions the court found insufficiently clear at the emergency stage. [6]

That distinction should not be flattened. A statutory-authority ruling says, in substance, that the agency lacked permission to impose the term. A vagueness ruling says the government did not give the recipient clear notice of what conduct would put funding at risk. Both theories protect recipients from surprise conditions, but they do different work. The first asks who had power to write the rule. The second asks whether the recipient could understand the rule before accepting federal money.

Judge Orrick’s July 10, 2026 preliminary injunction widened the pressure field. The Fresno-led coalition case, in which San Francisco participates, did not concern only one grant stream. It applied to grants from DHS, FEMA, Interior, and DOJ, and the ruling held that the challenged conditions likely violate separation of powers, the Spending Clause, and the APA. [7]

Orrick’s ruling is therefore important for portability, but still only at the preliminary-injunction stage. Its value for other recipients is the way it bundles three theories that often travel together in grant-condition challenges. Separation of powers addresses whether the executive branch is intruding into Congress’s spending role. The Spending Clause addresses whether Congress, or an authorized agency acting within Congress’s instructions, gave clear conditions on federal funds. The APA addresses whether the agency action can survive ordinary administrative-law review.

Why the DOE case belongs in the tracker but not in the pattern

San Francisco’s DOE challenge is useful because it shows the litigation line moving beyond police, school, and municipal emergency-management funds into clean-energy programming. The city says a new anti-DEI provision threatens approximately $130,000 in Clean Cities grant funding. [4][5]

But the case has no ruling yet. That makes it a live exposure item, not precedent. Counsel looking at analogous DOE or energy-transition grants should track the complaint and any temporary-relief motion, but should not cite the case as if a court has already accepted San Francisco’s statutory or constitutional theory. The most it currently shows is that the same grant-condition dispute has reached another agency and another program area.

Current risk posture for grant recipients

The early San Francisco rulings matter outside San Francisco because grant conditions usually arrive through standardized certifications, award terms, agency notices, and program guidance. A condition blocked in one plaintiff’s case may still appear in another recipient’s award package, especially before agency staff revise templates or field offices receive litigation instructions. That is the practical gap where local counsel operates: the order may be public, but the next compliance certification may still be due.

For now, the strongest recipient-side arguments appear where one of three facts is present: the authorizing statute affirmatively points toward diversity-related activity; the challenged term is too unclear to translate into compliance instructions; or the agency imposed the condition through an award term, certification, or guidance vehicle without a clear statutory hook. Those categories come from the rulings that exist, not from a final appellate settlement of the issue.

The incompleteness is just as important. The San Francisco v. DOJ trial set for August 2026 could become the first real merits inflection point in this cluster. [1] A trial ruling for San Francisco would give recipients something firmer than preliminary-relief reasoning. A trial ruling for DOJ, or a narrowed merits decision, would force a much more careful program-by-program read.

The OMB rulemaking is the other major moving part. On May 29, 2026, OMB proposed changes under 2 CFR Part 200, with comments due July 13, 2026, that could codify anti-DEI conditions across federal financial assistance. [1] If finalized in a legally durable form, that rule could change the statutory-authority analysis. Recipients would still have constitutional, APA, and program-specific arguments, but agencies would be less exposed to the simple objection that no generally applicable federal-assistance rule exists.

That does not mean the proposed rule would automatically cure every defect. A government-wide regulation cannot override a grant statute that points in the opposite direction, and it would still need to survive ordinary administrative-law and constitutional review. But it could make future briefing less clean than the early San Francisco cases, where courts could focus on agency-imposed conditions not found in congressional grant authorizations.

There is also a jurisdictional constraint that should not be treated as a footnote. The Ninth Circuit has held that some APA-based grant-termination claims may belong in the Court of Federal Claims under the Tucker Act, narrowing the district-court route for certain recipients challenging termination of funds. [1] That issue does not erase the San Francisco preliminary rulings, but it can affect where and how a recipient sues, especially when the requested relief looks like recovery of money rather than prospective restraint of an unlawful condition.

How far to carry the San Francisco rulings

The safest use of these cases is as a structured checklist for analogous grants, not as a blanket answer. Before advising a recipient to sign, refuse, sue, or reserve rights, counsel should identify the statutory authorization, the exact award condition, the agency vehicle that imposed it, the remedy needed before the next funding event, and whether the recipient is challenging a future condition or an actual termination.

• If the authorizing statute contains diversity-related language, Donato’s police-grant reasoning is the closest comparator.
• If the condition is impossible to operationalize, Chen’s Spending Clause vagueness analysis is the more useful starting point.
• If the condition appears across multiple agencies through standardized terms, Orrick’s separation-of-powers, Spending Clause, and APA framing may matter most.
• If the case concerns DOE Clean Cities or similar energy funding, the San Francisco DOE complaint is relevant as a live filing, but not yet as a ruling.
• If funds have already been terminated, venue and Tucker Act issues may become as important as the merits theory.

The national field is broader than these San Francisco-linked matters. Gibson Dunn’s June 8, 2026 update describes dozens of pending anti-DEI grant cases nationwide. [8] That broader universe may produce conflicting district-court rulings, appellate narrowing, or agency-specific distinctions. The San Francisco cases are influential because they are early, concrete, and clustered in one federal district; they are not the whole map.

The ACLS permanent injunction, entered on May 7, 2026 in the Southern District of New York, is a useful boundary marker because it is a final merits judgment, unlike most of the San Francisco-line rulings discussed here. [1] Its existence underscores the procedural caution rather than removing it: most of the San Francisco material remains preliminary.

San Francisco’s cases have created a coherent early body of law around anti-DEI federal grant terms. They have also left the decisive questions open. Until the DOJ trial, the DOE case, appellate treatment, Tucker Act routing issues, and the OMB rulemaking settle more of the terrain, grant recipients should treat this line as a developing risk map rather than settled doctrine.

References

1. San Francisco Finds Success Challenging Federal DEI-Related Grant Conditions, O’Melveny & Myers
2. Judge rejects DOJ attempt to tie police grants to anti-DEI pledge, Courthouse News
3. Judge Blocks DOJ Anti-Diversity Conditions On Police Grants, Law360
4. San Francisco says new anti-DEI provision threatens federal energy funds, Courthouse News
5. SF Sues DOE Over Clean Energy Grant's Anti-Diversity Terms, Law360
6. San Francisco Schools Win Early Block to Trump's DEI Grant Rules, Bloomberg Law
7. Judge stops new anti-DEI conditions for federal grants in California and Oregon, Courthouse News, July 10, 2026
8. Gibson Dunn June 8, 2026 update, Gibson Dunn, June 8, 2026