Purely Functional Computer Code Is Not Covered by the First Amendment Absent Plausible Allegations of Expressive Use

Defense Distributed; Second Amendment Foundation, Inc. v. Attorney General of New Jersey (3d Cir. Feb. 12, 2026)

I. Introduction

This Third Circuit decision arises from New Jersey’s efforts to curb the spread of “ghost gun” technology—particularly the distribution of digital files enabling 3D-printing of firearms and components. The plaintiffs, Defense Distributed (a publisher of “digital firearms information” or “DFI”) and the Second Amendment Foundation, Inc., challenged actions by the Attorney General of New Jersey and a state statute, N.J. Stat. Ann. § 2C:39-9(l)(1)-(2), that restrict distribution to New Jersey residents who are not registered or licensed as firearms manufacturers.

The case is notable not because it primarily turns on the Second Amendment, but because it forces a threshold First Amendment inquiry: when is computer code (including CAD/CAM-type files used in fabrication) “speech” at all? The Third Circuit affirms dismissal with prejudice, holding that First Amendment coverage for code is not automatic and depends on plausible, context-specific allegations of expressive purpose or use—allegations the complaint did not supply.

Key Issues

• Transfer/venue: whether law-of-the-case or comity required retransfer to the Western District of Texas after Fifth Circuit proceedings.
• Second Amendment standing: whether plaintiffs alleged a concrete injury from New Jersey’s manufacturing prohibition.
• Due process vagueness: whether “may be used” in § 2C:39-9(l)(2) fails fair notice or invites arbitrary enforcement.
• First Amendment coverage of code: whether the regulated files are covered by the First Amendment, and whether plaintiffs pleaded enough to decide that.

II. Summary of the Opinion

The court also affirms: (1) denial of retransfer to Texas; (2) dismissal of the Second Amendment claim for lack of standing; and (3) dismissal of the vagueness challenge.

III. Analysis

A. Precedents Cited

1. Transfer, law-of-the-case, and comity

• Arizona v. California, 460 U.S. 605 (1983): Used for the canonical description of law-of-the-case as a prudential doctrine promoting efficiency and consistency, but “not an inexorable command.” The Third Circuit emphasizes the doctrine presupposes a prior decision rendered in the same case by a court with jurisdiction over the matter.
• In re City of Phila. Litig., 158 F.3d 711 (3d Cir. 1998): Cited for the proposition that law-of-the-case governs a court’s exercise of discretion, rather than limiting power.
• Home Depot USA, Inc. v. Lafarge N. Am., Inc., 59 F.4th 55 (3d Cir. 2023) and Farina v. Nokia, Inc., 625 F.3d 97 (3d Cir. 2010): Support the principle that law-of-the-case extends only to issues “actually decided” in prior proceedings.
• Defense Distributed v. Bruck, 30 F.4th 414 (5th Cir. 2022) and Def. Distributed v. Platkin, 55 F.4th 486 (5th Cir. 2022): The Third Circuit treats these as procedurally informative but not binding. Crucially, the Fifth Circuit itself acknowledged it “lacks power” to order return after transfer, undermining any claim that its views became law-of-the-case in the Third Circuit action.
• Mast, Foos & Co. v. Stover Mfg. Co., 177 U.S. 485 (1900): Anchors the court’s comity analysis: comity “persuades; but it does not command.” The Third Circuit uses this to reject any argument that “requests” from another circuit/district are legally binding.
• Jumara v. State Farm Ins., 55 F.3d 873 (3d Cir. 1995): Provides the factor framework under 28 U.S.C. § 1404(a). The court credits the district court’s methodical balancing and finds no abuse of discretion.
• Danziger & De Llano, LLP v. Morgan Verkamp LLC, 948 F.3d 124 (3d Cir. 2020): Cited for abuse-of-discretion review of transfer rulings.
• Satterfield v. Dist. Att'y Phila., 872 F.3d 152 (3d Cir. 2017): Used to describe what constitutes an abuse of discretion (clearly erroneous facts, erroneous law, improper application).

2. Standing doctrine (Second Amendment claim)

• Spokeo, Inc. v. Robins, 578 U.S. 330 (2016) and Ellison v. Am. Bd. of Orthopaedic Surgery, 11 F.4th 200 (3d Cir. 2021): Provide the “concrete and particularized” and “actual or imminent” injury requirements.
• TransUnion LLC v. Ramirez, 594 U.S. 413 (2021): Emphasizes standing must be shown for each claim and form of relief.
• Warth v. Seldin, 422 U.S. 490 (1975) and Associated Builders & Contractors W. Pa. v. Cmty. Coll. of Allegheny Cnty., 81 F.4th 279 (3d Cir. 2023): Highlight that standing analysis often turns on the nature/source of the claim.
• FEC v. Cruz, 596 U.S. 289 (2022): Cited for taking merits allegations as true for standing, but still requiring a pleaded injury.
• Finkelman v. Nat'l Football League, 810 F.3d 187 (3d Cir. 2016): Used for the personal and individual nature of injury-in-fact.
• Friends of the Earth, Inc. v. Laidlaw Env't Servs. (TOC), Inc., 528 U.S. 167 (2000): Invoked in a footnote to show that associational standing would still require properly pleaded injury to a member.

3. Vagueness and due process

• FCC v. Fox Television Stations, Inc., 567 U.S. 239 (2012) and United States v. Williams, 553 U.S. 285 (2008): Provide the two-prong vagueness test (fair notice; arbitrary/discriminatory enforcement) and clarify that “perfect clarity” is not required. Williams is also central to the court’s rejection of “close cases” as vagueness problems.
• Ward v. Rock Against Racism, 491 U.S. 781 (1989): Quoted via Williams for the proposition that even speech-related regulations need not be perfectly precise.
• Posters 'N' Things, Ltd. v. United States, 511 U.S. 513 (1994): Cited to support that narrowing context and definitional specificity can provide “relatively clear guidelines” and reduce arbitrary enforcement risk.

4. First Amendment: coverage versus protection; expressive conduct; code cases

• Texas v. Johnson, 491 U.S. 397 (1989), United States v. O'Brien, 391 U.S. 367 (1968), and Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288 (1984): Used to frame the threshold “First Amendment even applies” inquiry, and to caution against labeling limitless conduct as “speech” based on subjective intent.
• City of Lakewood v. Plain Dealer Publ'g Co., 486 U.S. 750 (1988): Appears in the court’s discussion of Bernstein’s “nexus to expression” approach tied to censorship risk.
• Bernstein v. United States Department of Justice, 176 F.3d 1132 (9th Cir. 1999), reh'g granted, opinion withdrawn, 192 F.3d 1308 (9th Cir. 1999): Treated as “illuminating” despite withdrawal; supplies source vs. object code distinctions and an approach that examines actual expressive use in a technical field.
• Junger v. Daley, 209 F.3d 481 (6th Cir. 2000): Treated as taking a broader view—treating all source code as protected—an approach the Third Circuit declines to adopt categorically.
• Universal City Studios, Inc. v. Corley, 273 F.3d 429 (2d Cir. 2001) and Commodity Futures Trading Commission v. Vartuli, 228 F.3d 94 (2d Cir. 2000): Provide a taxonomy of “code communications” (to computer; to user; to programmer) and a strong emphasis on functional capability as nonspeech. The Third Circuit draws from Corley’s use-and-recipient sensitivity without adopting any one circuit’s formulation as comprehensive.
• Sorrell v. IMS Health Inc., 564 U.S. 552 (2011) and Bartnicki v. Vopper, 532 U.S. 514 (2001): Address “information” as speech in context. The Third Circuit distinguishes both and refuses to constitutionalize “information” in the abstract, especially where the regulated thing is a technical tool.
• Winter v. G.P. Putnam's Sons, 938 F.2d 1033 (9th Cir. 1991): Used as an analogy: some information artifacts (e.g., aeronautical charts) function like technical tools or “products,” not “pure thought and expression.” The court uses this to illustrate why “conveys information” does not end the First Amendment inquiry for code-like tools.

B. Legal Reasoning

1. The court’s central doctrinal move: separating coverage from protection for code

The opinion carefully treats “code is speech” as an overstatement. The Third Circuit adopts two linked propositions:

• Code can be covered by the First Amendment (i.e., it may fall within the Amendment’s domain when it functions as expression or communication).
• Coverage is not automatic, because code is inherently functional; therefore purely functional code with no expressive purpose, use, or intent is not covered.

This “coverage” framing matters procedurally: if plaintiffs cannot plausibly allege facts showing their code-related activity falls within First Amendment coverage, courts need not (and should not) proceed to protection questions like content-based scrutiny, prior restraint doctrine, or overbreadth.

2. A context-specific, factual inquiry—then a pleading-based stop sign

The court identifies several non-exhaustive factors relevant to whether a regulation burdens expressive code:

• Technical nature (e.g., source code vs. object code; and here, CAD/CAM and other file formats).
• How the code/file is used in context and how humans interact with it.
• Who is communicating to whom (human-to-machine, programmer-to-programmer, programmer-to-lay user).
• Purpose (function, expression, or a mix).
• What it communicates, if anything.

But the court does not apply these factors in a merits sense because the complaint fails at the gateway. Two pleading failures are decisive:

• Statutory fit is unclear: the statute targets files/code that “may be used to program” a 3D printer. The complaint lists varied file types and materials (including plain-text and PDFs) without clearly identifying which ones are CAD/CAM or otherwise within § 2C:39-9(l)(2).
• Expressiveness is conclusory: asserting that “each and every computer file” is “technical, scientific, artistic, and political” is treated as a “threadbare” legal conclusion under pleading standards (citing Ashcroft v. Iqbal, 556 U.S. 662 (2009) and Oakwood Lab'ys LLC v. Thanoo, 999 F.3d 892 (3d Cir. 2021)).

Importantly, the district court offered leave to amend with guidance; plaintiffs declined and stood on their complaint, supporting dismissal with prejudice.

3. Transfer and inter-circuit friction: comity as discretion, not compulsion

The opinion’s transfer section underscores institutional limits in multi-forum disputes. Even where another circuit has criticized a transfer, the receiving court is not bound by “requests” couched in mandamus or comity language, particularly where the originating circuit acknowledges it lacks power to compel return. The Third Circuit frames this as:

• Law-of-the-case cannot apply because the Fifth Circuit’s views were not issued in the “same case” over which it retained jurisdiction.
• Comity is a practice of persuasion; the New Jersey court acted within discretion by considering but not following the “request.”
• § 1404(a) balancing was thorough and not an abuse of discretion.

4. Standing and vagueness: conventional doctrine, strict application

On standing, the court requires a concrete allegation that plaintiffs (or members) attempted and were prevented from 3D printing—missing from the complaint. On vagueness, it reads “may be used” as tethered to a narrowly defined technical category (digital model instructions enabling printer programming), concluding the statutory “fact” is clear even if proof can be technical.

C. Impact

1. A meaningful Third Circuit precedent on code and the First Amendment

The decision establishes—at least within the Third Circuit—that litigants challenging regulation of code must do more than invoke slogans (“code is speech”). They must plausibly plead context showing the regulated code is being used (or is intended to be used) as expression in a way that implicates First Amendment interests. In practical terms, early dismissal risk rises for complaints that:

• fail to identify the specific file types and their functional relationship to regulated conduct,
• fail to explain the intended audience (programmers vs machines vs end-users) and the communicative content,
• rely on abstract assertions of “political” or “scientific” value without factual support.

2. Implications for “ghost gun file” litigation

For challenges to laws like § 2C:39-9(l)(2), this opinion suggests the decisive battleground may become:

• file specificity (what exactly is being restricted),
• functionality versus expressiveness (is it a blueprint/tool enabling fabrication, or communication for debate/education/critique),
• audience and use (sharing to enable printing versus sharing for discussion or research).

Plaintiffs may attempt to plead (and prove) “expressive use” by emphasizing educational, political advocacy, or scientific discourse functions—akin to Bernstein’s cryptographic peer-review narrative. Conversely, states may emphasize that certain file formats operate primarily as operational tooling for fabrication.

3. Spillover into other areas: cybersecurity, malware, AI models, and technical “tools”

While the court does not decide the constitutionality of regulating functional code generally, its coverage framework has obvious downstream relevance to disputes about: exploit code, ransomware, instructions enabling unlawful intrusion, and potentially the distribution of certain technical AI artifacts—where litigants may argue that a “tool” also “conveys information.” This opinion equips courts to demand a rigorous, use-sensitive account before elevating such disputes into heightened First Amendment review.

IV. Complex Concepts Simplified

• First Amendment “coverage” vs. “protection”: “Coverage” asks whether the First Amendment applies at all (is this speech/expression?). “Protection” asks what level of scrutiny applies and whether the regulation survives. The court stops at coverage because plaintiffs didn’t plead facts showing their files are within First Amendment coverage.
• Source code vs. object code: Source code is human-readable programming text; object code is machine-readable compiled code. Prior cases treated this distinction as potentially important to expressiveness. The Third Circuit treats technical form as a relevant factor, not a universal rule.
• CAD vs. CAM (in 3D printing): CAD often describes a design/model file; CAM is often closer to manufacturing-ready instructions. The statute targets files/code that can be used to program a 3D printer. The complaint’s failure to specify which file types were CAD/CAM or printer-programming-capable was critical.
• Void for vagueness: A law is unconstitutional if ordinary people can’t tell what it forbids, or if it gives police/prosecutors unguided discretion. Here, the court says the statute is narrow enough (specific digital model instructions enabling 3D-print programming) that it provides workable notice and enforceable facts.
• Comity: Respect among courts, not a binding command. A court may consider another court’s view, but need not “abdicate” its judgment.

V. Conclusion

The Third Circuit’s most significant contribution is its insistence that constitutional analysis of code begins with a disciplined threshold inquiry: code is not presumptively speech. Instead, First Amendment coverage depends on plausible, context-specific indicia of expressiveness; purely functional code with no expressive purpose, use, or intent falls outside the First Amendment’s coverage. Because the plaintiffs failed to plead facts allowing the court to determine what files were regulated and whether those files were expressive rather than operational, their First Amendment claims failed at the outset.

Separately, the opinion reinforces conventional doctrines in a high-profile setting: transfer disputes are governed by discretion (not inter-circuit command), Second Amendment claims require concrete injury for standing, and technical statutory language does not become vague merely because application may be complex.