O-1A Guide
O-1A Judging Criterion: A data scientist's Guide for November 2023
This guide covers the latest strategies and evidence requirements. Learn what changed and how to position your case.
The judging criterion and what's at stake for data scientists
The judging criterion is among the most practical O-1A criteria available to data scientists because the data science field generates substantial judging activity through peer review of research, technical program committee service, grant review panels, and machine learning competition evaluation. Unlike criteria such as receipt of a major prize or authorship of influential scholarly articles — which depend on past achievements that cannot be manufactured quickly — the judging criterion can often be satisfied through documented activities that an active data scientist may already have performed without recognizing their legal relevance to an O-1A petition. Understanding how to identify, document, and present this evidence is the first step in using the criterion effectively.
Satisfying the judging criterion is valuable not only as a standalone criterion but also because it corroborates other criteria. An invitation to review papers for a conference such as NeurIPS, ICML, ICLR, or ACL implicitly establishes that the organizing committee considered the applicant qualified to evaluate cutting-edge research in the field — a form of peer recognition that supports the broader extraordinary ability narrative. USCIS adjudicators are instructed to evaluate each criterion independently and also to assess the totality of evidence when no single piece of evidence is conclusive. Documented judging activity at distinguished venues strengthens the overall picture of the applicant's standing in the field.
Data scientists who have focused primarily on industry rather than academic research often assume they lack the peer review record necessary to satisfy the judging criterion. This assumption is frequently incorrect. Industry data scientists are regularly invited to review submissions to applied machine learning conferences, evaluate grant applications for programs such as NSF, NIH, or DARPA, serve on editorial boards for journals such as the Journal of Machine Learning Research or IEEE Transactions on Neural Networks, or judge Kaggle-adjacent competitions organized by institutions or professional societies. The key is to conduct a comprehensive audit of all judging activities before concluding that the criterion is unavailable.
What the regulation requires
The judging criterion at 8 C.F.R. § 214.2(o)(3)(iii)(B)(4) requires that the applicant has participated, either individually or on a panel, as a judge of the work of others in the same or an allied field of specialization. The regulation does not specify a minimum number of judging instances, a particular type of judging forum, or any formal credential required of the judge. The USCIS Policy Manual elaborates that the criterion can be satisfied by judging activities in the same or an allied field, which for data scientists can include computer science, statistics, artificial intelligence, engineering, and domain-specific fields such as bioinformatics, computational linguistics, or financial modeling depending on the applicant's area of expertise.
The phrase 'participated as a judge' encompasses a range of activities beyond formal competition judging. Peer review of academic papers submitted to conferences or journals is explicitly recognized as qualifying judging activity in USCIS policy guidance. Grant application review for government funding agencies — where a panel of experts evaluates proposals and makes recommendations on funding — also qualifies. Evaluation of competition submissions, review of manuscript drafts as an invited reviewer for academic publishers, and membership on technical program committees where committee members are responsible for making accept/reject recommendations on submitted papers are all forms of participation that can satisfy the criterion when properly documented.
The distinction of the judging forum matters to how USCIS weighs the evidence, even though the regulation does not enumerate specific qualifying venues. Reviewing papers for NeurIPS, ICML, ICLR, ACL, EMNLP, or ACM conferences — which have competitive submission processes, substantial international participation, and recognized standing in the machine learning and AI research community — is stronger evidence of expert-level recognition than reviewing for a regional conference with an open submission policy. USCIS adjudicators are not always familiar with field-specific venue hierarchies, so petitions should include context establishing why the conferences or journals at which the applicant has reviewed are considered authoritative in the data science and machine learning communities.
Evidence that satisfies the judging criterion for data scientists
The primary documentary evidence for peer review activity is the invitation from the conference program chair, journal editor, or grant review panel coordinator that specifically requested the applicant's participation as a reviewer. These invitations typically state the applicant's name, the conference or journal for which review is requested, the papers or manuscripts assigned for review, and the deadline. They should be retained and included as exhibits in the O-1A petition. Invitations received by email can be preserved as PDFs with full header information showing the sender's institutional affiliation and the date of the invitation.
In addition to the invitation, supporting documentation that establishes the distinction of the venue is essential. This can include the conference website showing acceptance rate data, the published proceedings demonstrating the competitive review process, the editorial board roster of the journal showing the academic standing of the editors and board members, or a statement from an expert in the field explaining the venue's standing in the data science community. For grant review panels at agencies such as NSF or NIH, the agency's public documentation of the grant review process and the selection criteria for panel members establishes both that the process is competitive and that reviewers are selected based on expertise.
Certificates or acknowledgment letters from conference organizers thanking the reviewer for their participation serve as corroborating documentation that the review activity was completed. Many major conferences publish lists of reviewers in their proceedings or on their websites — these public acknowledgments can be screenshotted and included as exhibits to demonstrate both that the applicant served as a reviewer and that the conference is a real, active organization with a documented community of participants. When multiple conference or journal review invitations are available, the petition should present the strongest three to five in detail and list additional instances in a table, prioritizing venues with the highest standing in the relevant specialty.
Evidence USCIS typically discounts
USCIS adjudicators have become increasingly attentive to the distinction between genuine peer review and superficial participation that does not reflect expert-level recognition. An invitation to serve as a reviewer sent as a bulk solicitation to all registered conference attendees does not demonstrate that the applicant was selected for their expertise — it demonstrates that they are a conference participant. Petitions that rely on review invitations from conferences that do not have competitive submission processes, that do not document how reviewers were selected, or that cannot demonstrate that the conference itself is a recognized venue in the applicant's specialty are at risk of adverse adjudication on the judging criterion even when the invitations are genuine.
Internal peer review within an organization — reviewing a colleague's code, evaluating a team member's analytical methodology, or assessing an internal research project — does not qualify as judging for O-1A purposes. The regulation and policy guidance contemplate external judging contexts in which the applicant's expertise is recognized by an independent organization and the applicant is asked to evaluate the work of individuals who are not the applicant's colleagues or subordinates. Internal technical review, however rigorous, is a standard job responsibility rather than an external recognition of expertise, and USCIS will not credit it as evidence satisfying the judging criterion.
Invitations to speak at conferences should not be characterized as judging activity. Serving as a conference speaker, panelist, or keynote presenter reflects recognition of the speaker's expertise but not participation as a judge of others' work. Similarly, writing a review of someone else's published book or article for a journal may constitute peer commentary but does not satisfy the regulatory criterion in the same way that pre-publication peer review of submitted manuscripts does. The petition should present judging evidence in the appropriate criterion category and should not characterize non-judging activities as judging to avoid both misrepresentation concerns and adverse adjudication when the adjudicator identifies the mischaracterization.
Framing borderline judging activity credibly
Kaggle competitions organized by companies or research institutions present a framing challenge: the competition format — where participants submit predictions to a leaderboard and are evaluated by scoring — is the reverse of what the regulation describes. Competition participants are evaluated rather than being evaluators. However, data scientists who serve as competition organizers, problem designers who define the evaluation rubric, or judges who evaluate final-stage submissions in competitions with a human-judgment component can potentially satisfy the criterion based on that organizational or evaluative role, provided the distinction of the competition is established and the applicant's specific judging function is documented.
Hackathon judging is more straightforward: data scientists invited to evaluate hackathon submissions and select winning projects are functioning as judges in the conventional sense, and this activity can satisfy the criterion when the hackathon itself is distinguished. The key is establishing distinction — a hackathon organized by a well-recognized technology company, a major university, or a prominent professional organization is more likely to be accepted as distinguished than one organized without clear institutional affiliation. The invitation to judge, the criteria used for evaluation, and the outcome of the competition all support the documentation of a genuine judging role.
For data scientists who have served on academic dissertation or thesis committees — evaluating whether a graduate student's research meets the standard for a degree award — this activity constitutes judging in an allied academic field and can be documented with appointment letters from the university department and the thesis defense record showing the applicant's participation as an examiner. Dissertation committee service at a recognized research university is a form of expert evaluation that USCIS has accepted as satisfying the judging criterion in prior adjudications. The petition should include documentation of the university's standing, the research area of the dissertation, and the applicant's role as an external expert evaluator rather than as the student's advisor or supervisor.
An audit checklist for the judging criterion
Before filing, data scientists should conduct a systematic inventory of all potential judging activities across their career. The inventory should cover: peer review invitations for academic conferences in data science, machine learning, AI, and allied fields; manuscript review invitations from journals including IEEE Transactions on Neural Networks and Learning Systems, Journal of Machine Learning Research, Artificial Intelligence, and domain-specific journals; grant review panel participation for NSF, NIH, DARPA, or equivalent funding agencies; competition judging roles at hackathons, machine learning challenges, or industry competitions with institutional organizers; dissertation or thesis committee appointments; and editorial board memberships at academic publications. Documentation for each activity should be located or reconstructed before the attorney drafts the petition.
For each judging activity that will be presented in the petition, verify that the following documentation is available: the invitation from the organizing body, which should state that the applicant was selected to serve as a reviewer or judge and ideally explain the basis for the selection; evidence of the venue's distinction, such as acceptance rate data, proceedings, editorial board composition, or sponsoring institution information; and confirmation that the review was completed, such as an acknowledgment from the organizer, inclusion in a published reviewer list, or correspondence about the review outcome. Where documentation gaps exist, the attorney should determine whether the missing evidence can be reconstructed through archived emails, conference websites, or communication with the organizing body.
The petition's judging criterion section should present evidence cumulatively rather than relying on any single judging instance to carry the criterion alone. Even a strong single judging invitation can be dismissed as an isolated event; multiple documented review activities across different venues and time periods establish a pattern of expert recognition that is more persuasive as a criterion demonstration. The cover letter should identify each piece of judging evidence by exhibit number, establish the distinction of each venue or program, and draw the connection to the regulatory language — that the applicant has participated as a judge of the work of others in the same or allied field — explicitly rather than leaving that analysis for the adjudicator to perform independently.