David Kozik, a Senior Executive Service official in the Office of the Undersecretary of Defense for Intelligence and Security, signed his name to a document that the Pentagon would rather most people never read.
The document — released through a Freedom of Information Act request and catalogued as 23-F-0241 — is a formal Defense Department appeal against a piece of House legislation that would have required the Pentagon to build a dedicated, publicly advertised system for personnel to report unidentified aerial phenomena, even if doing so meant disclosing information covered by a non-disclosure agreement. The appeal was reviewed by an attorney in the Office of General Counsel — Intelligence. The language is bureaucratic. The intent is not.
The appeal was reviewed by an attorney in the Office of General Counsel — Intelligence.
The Pentagon wanted the provision killed.
What the House actually proposed
Section 1663 of H.R. 7900, the House version of the fiscal year 2023 National Defense Authorization Act, was a straightforward attempt to solve a problem that UAP researchers and some lawmakers had identified for years: personnel who believed they had witnessed or had knowledge of UAP-related programs had no safe, formal channel to report what they knew without risking prosecution under secrecy agreements. The House provision would have changed that.
Under the proposal, the Pentagon would have been required to establish a secure reporting system, administered by employees of the newly created All-domain Anomaly Resolution Office — AARO — whose identities and contact information would be publicly posted on a DoD website. Any federal employee or contractor could use the system to report UAP-related events or programs, even if that information was covered by a written or oral non-disclosure agreement. Critically, the heads of all federal departments and agencies that had ever supported UAP investigations would be required to search comprehensively for records relating to NDAs and related obligations, and those records would be made accessible to the congressional intelligence committees, the congressional defense committees, and congressional leadership.
The Senate Armed Services Committee included no similar provision in its version of the NDAA. The Senate Select Committee on Intelligence did include a comparable measure in its mark of the Intelligence Authorization Act for Fiscal Year 2023 — S. 4503, section 704 — but the House provision was the more expansive of the two. That asymmetry mattered: it meant the House language would have to survive a conference negotiation against a Senate counterpart that was, at minimum, less aggressive.
The Pentagon moved to ensure it didn't survive at all.
The objections, and what they reveal
The 23-F-0241 appeal laid out the department's objections in careful, lawyered prose. Read in sequence, they trace a hierarchy of concern that tells you something about institutional priorities.
The first and most prominent objection was constitutional: the House provision, the Pentagon argued, could intrude on the president's authority to manage classified information and information security systems. That argument was broad enough to apply to almost any independent oversight mechanism involving classified material — which is presumably the point.
The second objection was security-based. Requiring a publicly acknowledged system, with clear guidance posted online about how to access and use it, would create unnecessary national security and counterintelligence risks. The department argued that any adversary could study the public-facing system and learn something useful about how classified UAP information was being handled.
The third objection was operational. Building and managing the system would be resource-intensive and would divert personnel and money from other AARO functions. The document stated plainly that any statutory requirement to establish a new system outside of existing mechanisms would require the expenditure of personnel and financial resources that would be better served elsewhere.
The fourth objection was more specific — and more revealing. The requirement that the system be administered by employees described as widely known and easily accessible, the Pentagon said, could expose those personnel to privacy, security, and personal safety risks, given what the document called the potentially polarizing nature of the UAP subject. The department said it would object specifically to the inclusion of those words in subsection (b)(3) of the provision.
The fifth objection addressed the NDA search requirement: the Pentagon said the language assumed the department had the resources and capability to search for, retrieve, and provide all records relating to non-disclosure agreements or orders — some of which may be decades old and implicate the Privacy Act — to satisfy what it characterized as an unspecified congressional purpose.
Then came the sentence that deserves to be read twice. The document stated that the department shares the goal of establishing a secure mechanism to encourage and protect individuals who provide potentially classified information regarding UAP or UAP-related programs to authorized personnel in the AARO. But it immediately followed that declaration by arguing that existing authorities — specifically Executive Order 13526 and Executive Order 12968, governing classified information and personnel security — already provided such a mechanism, and that a new statutory system was therefore unnecessary.
This is where the document's internal tension becomes most visible. The Pentagon claimed to share Congress's goal while simultaneously arguing that the goal had already been achieved through existing mechanisms. If that were true, the House provision would have been largely redundant — a paperwork burden, nothing more. But the vehemence of the department's opposition, the breadth of its objections, and the formal nature of the appeal all suggest something else was at stake.
The document urged outright exclusion of the House provision. Failing that, it said the department was eager to work with conferees to revise it.
The provision did not survive the conference in the form the House passed it.
The gap that the document opens
What 23-F-0241 does not contain is equally important. There is no accompanying record of what the Pentagon proposed as an alternative. There is no documentation of what existing mechanisms — if any — were being used to receive UAP-related disclosures from personnel who believed they were bound by NDAs. There is no response from congressional staff explaining why the House language was written the way it was, or what problem the drafters believed existing executive order authorities had failed to solve.
The document's provenance adds another layer. It was released via FOIA — meaning someone requested it, the department processed it, and it entered the public record through that specific channel. What was redacted, and why, is not fully visible in the released version. The submitting component is redacted. The reviewing attorney's name is redacted. Only Kozik's name and title remain legible in the internal control section.
The Senate Select Committee on Intelligence, meanwhile, included its own analogous provision in S. 4503. Whether that provision survived the intelligence authorization process — and in what form — is a separate question, and one the document does not answer.
None of this is, by itself, proof of bad faith. Bureaucratic resistance to new oversight mechanisms is common, and the Pentagon's procedural and security objections are not inherently implausible. Large institutions routinely defend existing authorities against statutory encroachment. That is normal.
What is not normal — or at least not unremarkable — is the specific context. The House provision was written in response to years of claims by researchers, former government officials, and at least some members of Congress that personnel with UAP-related knowledge had been silenced by classification and NDA obligations. The Pentagon's response was to argue that the system for handling such disclosures already worked fine, without providing any evidence that it had ever been used for that purpose.
The claim that existing mechanisms were sufficient and the institutional effort to block a new independent channel are not necessarily contradictory. But they sit in uncomfortable proximity to each other — and the department has not, in any publicly available record, reconciled them.
The specific document or congressional record that would most change this picture is a detailed accounting of how the AARO actually handled UAP disclosures under the existing authorities the Pentagon cited in its appeal — how many reports came in, under what classification levels, and whether any involved personnel who had previously signed non-disclosure agreements. That accounting, if it exists, has not been made public.