O-1A Guide
O-1A for AI Ethics Researchers: NSF Ethics in AI Grants, AI and Society Publications, and Field Recognition in 2026
AI ethics researchers publish across computer science conferences, philosophy journals, and law reviews, creating a nuanced scholarly articles criterion challenge. Understanding which publication types satisfy the O-1A regulatory standard and how to present borderline venues including ACM FAccT and policy reports is essential to building a strong petition.
The scholarly articles criterion in AI ethics research
The scholarly articles criterion under 8 C.F.R. § 214.2(o)(3)(iii)(B)(6) requires evidence of the petitioner's authorship of scholarly articles in the field, in professional journals or other major media. For AI ethics researchers — scholars and scientists who study the social, ethical, legal, and technical dimensions of artificial intelligence systems — this criterion is often the strongest anchor for an O-1A petition. The field produces a distinctive mix of peer-reviewed journal publications, conference papers, policy reports, and public-facing essays, not all of which qualify equally as scholarly articles in professional journals or other major media. Understanding which publication types satisfy the criterion and how to present them is central to building a strong scholarly articles case in this field.
AI ethics research spans computer science, philosophy, law, sociology, psychology, and public policy. The diversity of disciplinary homes creates a complex publication landscape where the same researcher may publish in IEEE conference proceedings, philosophy journals, law reviews, and policy institute reports over the course of a year. USCIS does not restrict scholarly articles to any particular discipline — publications in peer-reviewed philosophy journals count alongside conference papers in computer science venues — but the petition must explain the significance of each publication type to an adjudicator who may not be familiar with how different academic disciplines evaluate research output. The cover letter and expert letters must translate disciplinary conventions into language that allows the adjudicator to apply the regulatory standard accurately.
NSF's Ethics and Values in Science and Technology program and the Responsible Computing programs within the Directorate for Computer and Information Science and Engineering provide competitive grant funding for AI ethics research. An NSF award in these programs reflects merit review by a panel of researchers who evaluated the proposed research against NSF's standard of intellectual merit and broader impacts. For an AI ethics researcher, NSF grant funding supports the scholarly articles criterion indirectly by demonstrating that independent experts have judged the proposed research program — the program that will produce publications — as significant enough to fund. The establishment of dedicated NSF funding mechanisms for AI ethics reflects the field's recognized importance and can be cited in the petition to contextualize the petitioner's research within federally supported research priorities.
What the regulation requires
The regulatory text specifies publications in professional journals or other major media. Professional journals refers to peer-reviewed journals that publish original research for expert audiences in a recognized academic or professional discipline. Other major media is a broader category that has been interpreted to include law reviews, technical magazines with peer review, and academic book chapters in edited scholarly volumes. For AI ethics researchers, the key distinction is between publications that have undergone peer review or professional editorial selection by recognized experts, and publications that were produced through the petitioner's own platform or through venues without expert gatekeeping — blog posts, newsletter essays, and organizational reports authored entirely by the petitioner without external review.
The peer review standard does not require anonymous peer review in every case, but it does require that publication decisions be made by independent experts in the relevant field rather than solely by the petitioner or the petitioner's employer. A policy report published by a major research institution involves expert authorship and editorial review, but whether it qualifies as a scholarly article in professional publications depends on the publication's standing in the relevant academic community and the review process it applies. The petition should document the publication venue's selection criteria and editorial process for each publication cited, particularly for venues that are not traditional academic journals with well-known peer review processes.
Conference proceedings papers in computer science and AI ethics occupy a distinctive position in the publication landscape. In computer science, conference papers published in the proceedings of major venues including ACM FAccT, AAAI, NeurIPS, ICML, and CHI undergo peer review and are considered primary scholarly publications. ACM FAccT papers on fairness, accountability, and transparency are directly relevant to AI ethics research and qualify as scholarly articles in a venue with an established competitive peer review process. The petition should document the acceptance rate and review process for each conference venue cited, because acceptance rates at top AI venues — typically well under 20 percent — provide evidence of the competitiveness of the selection process through which the petitioner's work was accepted and evaluated.
Evidence that routinely satisfies the criterion
Peer-reviewed journal publications in AI and Society, Ethics and Information Technology, Science and Engineering Ethics, the Journal of Artificial Intelligence and Law, and Philosophy and Technology represent core scholarly publication venues for AI ethics research. These journals have established editorial boards composed of recognized experts, undergo formal peer review processes, and publish original research that is cited by subsequent scholars in the field. First-author publications in these journals, presented with impact factor, acceptance rate where available, and citation count, provide straightforward documentary evidence for the scholarly articles criterion without requiring extensive additional explanation of the publication venue's significance.
Publications in top interdisciplinary or general academic journals represent the strongest evidence for the scholarly articles criterion. A paper in Nature Machine Intelligence, Science, PNAS, or Nature Communications on an AI ethics topic — algorithmic bias in high-stakes decision systems, societal impacts of large language models, or governance frameworks for autonomous systems — demonstrates that the petitioner's work was evaluated through a highly competitive peer review process that accepts a small fraction of submissions. These publications should be presented with full citation information, journal acceptance rate or impact factor, and a brief expert annotation explaining the significance of the research finding and the reception the paper has received in the field.
Cross-disciplinary publications in law reviews and policy journals provide evidence that the petitioner's AI ethics research has been recognized outside computer science. Law reviews at major research universities are among the most selective academic publications in their respective disciplines, and an AI ethics paper published in such a venue demonstrates both the scholarly quality of the work and its interdisciplinary significance. Policy institute publications with rigorous standards — Brookings Institution research reports, RAND Corporation technical publications, or reports published through the National Academies of Sciences involving external peer review — also provide scholarly article evidence when the publication process involves expert review and attributes the work clearly to the petitioner as an author.
Evidence USCIS regularly discounts
Blog posts, newsletter essays, and social media threads — regardless of how widely they are read or shared — do not satisfy the scholarly articles criterion because they are not published in professional journals or other major media as those terms have been interpreted by USCIS. An AI ethics researcher who has published influential essays on personal or professional platforms may have significant public impact, but that impact reflects influence in public discourse rather than recognition by the professional scholarly community through structured peer review. These materials can be included in the petition as supporting evidence for other criteria — particularly press and media coverage or original contributions — but should not be presented as satisfying the scholarly articles criterion.
Reports authored entirely by the petitioner without external peer review do not satisfy the scholarly articles criterion. Internal organizational research reports, self-published white papers, or reports produced solely by the petitioner's research group lack the external expert validation that the criterion requires. Similarly, conference workshop papers and extended abstracts that have not undergone full paper review equivalent to a peer-reviewed proceedings paper do not provide strong scholarly articles evidence. The criterion is specifically about external expert validation of the petitioner's work through publication selection by independent reviewers, and materials that lack this external validation element should be presented under other criteria where they may be more persuasive.
Co-authored reports from think tanks or advocacy organizations, even when co-authors include recognized experts, may receive limited weight for the scholarly articles criterion if the publication venue does not have an established peer review process comparable to an academic journal. Many AI ethics research outputs are produced as policy-facing reports by organizations such as the AI Now Institute, the Center for Democracy and Technology, or the Future of Life Institute, and these reports are often influential and widely cited. When a petition presents these as evidence for the scholarly articles criterion, it should document the review process the organization applies to publications and distinguish between processes with meaningful external review and those that rely primarily on internal review by the organization's own staff.
How to present borderline evidence
Conference papers at top AI venues occupy an ambiguous space that the petition should address directly. ACM FAccT, NeurIPS, ICML, ICLR, ACL, and similar venues have acceptance rates that make them as selective as leading journals in many research fields. Some USCIS adjudicators and RFE templates have questioned whether conference papers qualify as publications in professional journals under the regulation. The petition should address this directly by documenting that computer science conference papers are the primary form of scholarly publication in the field, that papers at these venues undergo anonymous peer review by independent experts, and that citation practices in computer science treat conference papers as equivalent in significance to journal articles. Expert letters from computer scientists explaining this disciplinary norm are essential for this argument.
Law review articles by AI ethics researchers present a different kind of borderline case. Law review submission and publication practices vary by journal — some journals use entirely student editors without expert peer review in AI ethics or computer science, while others employ external expert reviewers for technically complex submissions. A law review article selected and edited without review by AI ethics or computer science experts may be a significant legal publication but does not demonstrate extraordinary ability in AI ethics research specifically. The petition should explain for each law review publication cited whether the submission was reviewed by experts in the relevant technical field, and calibrate the claimed significance of the publication to the nature of the review process it underwent.
Preprints posted to arXiv or SSRN before peer-reviewed publication should be distinguished from published papers in the petition. A preprint that was subsequently published in a peer-reviewed journal is properly cited by reference to the final published version; citing both the preprint and the published version as separate publications would mischaracterize the record. A preprint that has not yet been published in a peer-reviewed venue does not satisfy the scholarly articles criterion on its own, though it may be cited as evidence of ongoing research productivity or as context for understanding a subsequent publication. The petition should present only published, peer-reviewed work as primary scholarly articles evidence.
Building and auditing the scholarly articles file
The scholarly articles file should be organized to present the strongest publications first and to explain each publication's significance in accessible terms. The exhibit should include a publication list identifying each paper's title, venue, year, author position, and a brief description of the contribution, followed by individual exhibits for each publication showing the title page with the petitioner's authorship, the venue's name and impact factor or acceptance rate, and a citation count. For publications in specialized AI ethics journals or interdisciplinary venues, a one-paragraph expert annotation explaining the venue's significance and the paper's reception in the field helps the adjudicator understand why the publication represents evidence of extraordinary ability.
Before finalizing the scholarly articles exhibit, the petitioner's representative should run a self-audit against three questions: Is each publication in a venue that involves external expert review? Has the petitioner's authorship position been clearly documented? And has the significance of each publication been explained so that an adjudicator without specialized knowledge of AI ethics research can understand why it constitutes scholarly contribution at a level consistent with extraordinary ability? If the answer to any of these questions is no for any publication in the exhibit, that entry needs revision or supplemental explanation. A petition that forces the adjudicator to infer the significance of publications without guidance is less persuasive than one that makes the argument explicitly.
The scholarly articles criterion interacts productively with the original contributions and press coverage criteria in an AI ethics petition. Publications that document original contributions — a novel fairness metric, a new method for auditing algorithmic decision systems, or a governance framework that has been adopted in policy discussions — satisfy both the scholarly articles and original contributions criteria simultaneously. Press coverage about the petitioner's published research in mainstream technology media such as The New York Times, Wired, or MIT Technology Review provides independent evidence that the petitioner's scholarly contributions have been recognized beyond the academic community. A petition that draws these connections — showing that the same research has been peer-reviewed for publication, widely cited, and covered by major media — presents a coherent and mutually reinforcing case for extraordinary ability in the field.
What we typically gather for this kind of case
| Document | Where to source | Why it matters |
|---|---|---|
| Peer-reviewed publications | Web of Science / Scopus exports | Anchors original-contributions and authorship criteria |
| Citation analysis | Google Scholar profile + ESI top-1% data | Quantifies major significance in the field |
| Salary benchmark | BLS OEWS for SOC code + locality | Documents high-salary criterion at 90th-percentile or above |
| Critical-role letters | Direct supervisor + program director | Establishes role's importance, not just title |
What we see go wrong, again and again
- 01Treating extraordinary ability as a credentials checklist rather than a story of field-wide impact.
- 02Submitting bibliometric data (h-index, citation counts) without explaining what makes those numbers high relative to peers in the same sub-field.
- 03Relying on letters from collaborators or co-authors rather than independent experts who can speak to influence.
See if you qualify
Lando reviews your background against the O-1A visa criteria and tells you honestly where you stand. Free, no commitment.