ATF’s New Rules Would Make Gun Records Searchable. That’s the Real Registry Problem.

The quiet shift from decentralized paper forms to structured, queryable electronic data removes the deliberate friction that historically prevented a centralized federal firearms registry. (Image generated via AI)

(The following article was originally published by the University of Wyoming’s Firearms Research Center on May 28, 2026.)

On April 29, 2026, the DOJ and ATF released thirty-four final and proposed rules under Executive Order 14206, “Protecting Second Amendment Rights.” The package represents a genuine achievement and step in the right direction, with several components delivering relief that gun-rights advocates have sought for years. Three of the proposed rules, grouped under the heading “ Modernize,” deal with the federal recordkeeping system: the Form 4473, the acquisition and disposition records that dealers keep, and record retention periods. The public response to the recordkeeping rules has in large part focused on the retention proposal, specifically on a question within it: how long the government may keep a transaction record. That focus is understandable, as retention duration is the most legible term in the rules. It is also, this article contends, the wrong variable to focus on. When the federal prohibition on a firearms registry was drafted it did not only focus on how long records were kept. It was also drafted to control how those records were organized and retrieved, and in that respect, the modernization rules move the system in a direction that has not yet been fully examined.

The Package, Fairly Stated

Any fair assessment of the package should start with conceding what it accomplishes. The agency proposes to rescind the 2023 stabilizing-brace rule, remove the bump-stock language from the machine-gun definition in conformity with *Garland v. Cargill*, clarifies that ordinary incidents of interstate travel remain protected under the Firearm Owners’ Protection Act’s transport provision, eliminates the chief law enforcement officer notification requirement for National Firearms Act filings, and authorizes married couples to register NFA firearms jointly without first creating a trust. Concerning the main operational gains: the agency’s own processing data show individual eForm 4 transfers clearing in a median of six days, compared to the many months that were until recently routine. These are major changes. The analysis argues for completing this reform.

Friction and Decentralization by Design

The shorthand term “no registry” understates how old and specific the federal ban is. Section 926(a) of Title 18, provides that no rule or regulation issued after the Act’s enactment:

“may require that records required to be maintained under this chapter or any portion of the contents of such records, be recorded at or transferred to a facility owned, managed, or controlled by the United States . . . nor that any system of registration of firearms, firearms owners, or firearms transactions or dispositions be established.”

The sentence contains two prohibitions. The first is centralization, or the physical or electronic gathering of records at a government facility. The second is registration, or the establishment of a system that associates owners with the firearms they own. Commentary on the retention rule has concentrated primarily on the second clause. The first has gone largely unremarked, but it is the more important one.

Read alongside the recordkeeping and tracing provisions that accompany it, the prohibition should be seen as Congress authorizing one narrow use of dealer records and categorically excluding all other uses. Section 923(g)(4) requires a licensee who goes out of business to deliver its records to the Attorney General, the only channel that closed dealer records have to lawfully reach federal custody. Section 923(g)(7) permits the agency to demand trace information, but only “for determining the disposition of one or more firearms in the course of a bona fide criminal investigation,” and requires a response within 24 hours. And since 2012, an appropriations rider, made permanent by its own word of futurity, “hereafter,” has barred the agency from using funds “to electronically retrieve information gathered pursuant to 18 U.S.C. § 923(g)(4) by name or any personal identification code.”

The guiding principle across all of these provisions is not the lifespan of a record, but its ability to be retrieved. Congress allowed the government to hold out-of-business records; however, it banned the government from organizing those records into something that answers the question about who owned what. The serial-number trace it authorized runs in one direction, from a firearm to a first purchaser, one inquiry at a time. The name-indexed retrieval it prohibits runs in the other direction. The practical difficulty associated with converting a warehouse of decades old transaction records into a digital, queryable index of gun owners is not an inefficiency Congress failed to consider, it’s the mechanism on which the law depends.

That reframing is important as it identifies the key legal variable the registry prohibition governs: not the duration of custody, but the structure of the records that are held. The point has acquired new importance since *Loper Bright Enterprises v. Raimondo*. A reviewing court will now construe “system of registration” and “consolidating or centralizing” using its own independent judgment, without deference to the agency’s narrower readings. The leading precedent permitting record retention against a registry challenge, , upheld the FBI’s NICS audit log under Chevron and on the grounds that the Brady Act’s command to destroy records did not specify when. That reasoning was a product of a deference regime that no longer exists. Whether a standardized, name-sortable digital recordkeeping system is a “system of registration” is now an open question of statutory meaning, and it is the unspoken question the modernization rules quietly raise.

From Image to Index

The three “Modernize” proposals reorganize the recordkeeping environment along the axis just identified. The Form 4473 proposal (RIN 1140-AA82) would authorize electronic forms, auto-population of fields, and the attachment of digital copies of identification documents in place of hand-transcribed entries. The product is a transaction record that starts as structured digital data versus a scanned image of a paper form.

The companion electronic-recordkeeping proposal (RIN 1140-AA94) is the more important of the two, as the rule would authorize licensees to generate, maintain, and store Forms 4473 and acquisition and disposition records electronically, and would fold several ATF rulings, including Ruling 2022–01, that governs electronic 4473 storage, into the Code of Federal Regulations. In laying out what a compliant electronic system must do, the proposed rule says records must be

“searchable by key terms including transferee name,” and “sortable . . . alphabetically (by purchaser name).”

That is ATF, in the rule now open for comment, describing a recordkeeping architecture indexed by the name of the purchaser.

In fairness to ATF the proposal does not discard all safeguards. When electronic records are eventually transferred to the Out-of-Business Records Center, ATF must strip optical character recognition be disabled and flatten images so federal copies are not text-searchable. But that safeguard should be viewed as a mere configuration instruction, applied at the moment of transfer, and existing downstream of a system the same rulemaking would standardize as name-sortable at the point of creation. The friction on which Section 926(a) and the appropriations rider rely on is not being defied; it is being relocated. The rules engineer it out at the licensee’s premises, where the records are created, and preserve it in attenuated form at the point of federal intake. Once the records that dealers create are uniformly born-digital and name-indexed, the durability of the entire anti-registry compromise comes to rest on whether one image-flattening step is reliably performed and never quietly bypassed or rescinded.

The natural objection is that the electronic-recordkeeping rule does not change anything, because ATF has allowed electronic 4473 storage by variance for years. That objection understates what codification does. A variance is optional and individually granted; a codified rule makes electronic, name-sortable recordkeeping the industry default and steadily increases the share of out-of-business records that reach federal custody already structured for retrieval. Codification is also the lawmaking moment, or the point at which the architecture is fixed in regulation. If the rule does no more than restate existing practice, then writing an explicit architectural limit into its text costs the agency nothing. This raises another compounding concern. As my recent Working Paper argued, once firearms records exist as standardized digital data, using AI, registry-equivalent knowledge could be derived by inference, without any formally declared registry ever being built. The modernization rules shorten the distance to that outcome.

Broadening the Comment Record

None of this counsels against the reform package as a whole, and the scrutiny the retention rule has drawn is deserved. The point here is additive. The retention proposal establishes the duration records should remain; the Form 4473 and electronic-recordkeeping proposals determine how those records are structured while they exist, and Section 926(a) targets structure, not duration. Comments and focus directed mostly at the retention period would leave the more important rules underexamined.

A focused recommendation follows from the statute. Commenters on RIN 1140-AA82 and RIN 1140-AA94 should ask ATF to place the architectural commitment in the text of the rules themselves: electronic records transferred to the agency under Section 923(g)(4) should be stored and retrievable only in a manner consistent with Section 926(a) and the appropriations rider, not indexed or retrievable by name or personal identification code, and that should be a regulatory requirement and not just an inspection-format instruction that a future administration can revise or rescind. The modernization of federal firearms recordkeeping could be consistent with Section 926(a) and should proceed. But precisely because this package reorganizes the data environment, it is the key moment to integrate the registry prohibition directly into the architecture rather than to rely on a paper-era friction the package removes. Public comment on all three “Modernize” proposals ends in early August 2026. The window in which to make the architectural point is the same window in which the architecture is being decided.

Originally published at https://firearmsresearchcenter.org on May 28, 2026.

Del Schlangen is an independent researcher writing at the intersection of artificial intelligence, emerging technology, law, and national strategy. You can follow his work here or connect with him on Reddit , X, or via email