NIW for AI Researchers
Artificial intelligence researchers may qualify for an EB2 National Interest Waiver when their specific proposed endeavor satisfies the Matter of Dhanasar framework.
Artificial intelligence can involve technologies important to U.S. competitiveness and national security, but simply working in AI does not automatically establish NIW eligibility. USCIS evaluates the particular proposed endeavor and the complete evidence.
For the broader computer science framework, review NIW for computer science researchers.
Underlying EB2 Eligibility
An AI researcher must first qualify for EB2 as an advanced-degree professional or individual of exceptional ability.
A master’s degree, PhD or other advanced STEM education may establish part of that threshold, but the degree alone does not establish the national interest waiver.
The Three Dhanasar Prongs
The applicant must establish:
- substantial merit and national importance of the proposed AI endeavor;
- that the applicant is well positioned to advance the endeavor; and
- that, on balance, waiving the job offer and labor certification would benefit the United States.
Define the AI Endeavor
The proposed endeavor should identify the specific AI problem and research direction.
A statement such as “I intend to work in artificial intelligence” is generally much less informative than a detailed endeavor involving a particular technical objective.
The description may identify:
- the AI method or model;
- the technical problem;
- the application domain;
- the research methodology;
- safety, performance or reliability objectives;
- anticipated users or industries; and
- the prospective broader impact.
Examples of AI Research Areas
Depending on the specific work, AI research may involve:
- machine learning;
- foundation models;
- computer vision;
- natural language processing;
- robotics;
- autonomous systems;
- AI safety;
- AI security;
- medical AI;
- scientific machine learning;
- AI hardware; or
- other specialized areas.
Critical and Emerging Technology
USCIS guidance recognizes that certain areas of artificial intelligence may qualify as critical and emerging technologies relevant to U.S. national security and competitiveness.
This can support the analysis, but it does not eliminate the requirement to evaluate the applicant’s specific endeavor.
National Importance
An AI endeavor may have national importance where the evidence demonstrates potential broader implications for areas such as:
- national security;
- economic competitiveness;
- health care;
- scientific discovery;
- advanced manufacturing;
- cybersecurity;
- energy;
- transportation;
- communications;
- education; or
- critical infrastructure.
The petition should demonstrate the connection rather than merely list nationally important industries.
Well Positioned
Evidence may include:
- advanced AI or computer science education;
- research publications;
- citations;
- AI models or algorithms;
- open-source software;
- datasets;
- patents;
- research grants;
- peer review;
- technical competitions;
- industry deployment;
- government interest;
- collaborations; and
- prior research success.
An Advanced STEM Degree Is Not Enough
USCIS considers advanced STEM education relevant to the well-positioned analysis, particularly for endeavors involving critical and emerging technology.
However, the degree by itself is not sufficient to establish that the applicant is well positioned to advance the endeavor.
AI Publications
Peer-reviewed AI research may demonstrate expertise and prior success.
Conference publications can be particularly important in some AI specialties, and the petition may explain field-specific publication practices.
Citations
There is no fixed NIW citation threshold for AI researchers.
Citations can help demonstrate that other researchers have noticed or used the applicant’s work.
Review NIW citation requirements.
Open-Source AI Projects
Open-source models, libraries and research tools can provide evidence of technical impact where the record demonstrates meaningful independent use.
Potential evidence includes:
- downloads;
- repository activity;
- independent adoption;
- integration into other software;
- research citations; and
- use by companies or institutions.
Datasets and Benchmarks
An AI researcher may develop datasets or benchmarks used by other research groups.
Usage data, academic citations and independent implementation may demonstrate influence.
Industry AI Researchers
Researchers employed by technology companies may possess evidence of large-scale practical impact that is not fully reflected by publication counts.
Examples include:
- production deployment;
- commercial adoption;
- patented models or methods;
- performance improvements;
- AI safety improvements;
- technical infrastructure;
- enterprise implementation; and
- measurable use of the technology.
AI Safety and Security
Research concerning reliable, secure or safe AI systems may have broader implications where the proposed work affects deployment of AI in significant industries or governmental functions.
The petition should document the specific technical research and its prospective impact.
Government Interest
Government funding, collaboration or documented agency interest may provide strong evidence when it relates directly to the applicant’s proposed AI endeavor.
Recommendation Letters
Independent AI researchers or technical experts can explain why the applicant’s work is important and how the applicant is positioned to continue advancing it.
Letters should contain specific technical facts and should be corroborated where possible.
The Third Dhanasar Prong
Even when an AI endeavor is important and the applicant is well positioned, the petition must separately address why granting the waiver would benefit the United States on balance.
NIW Versus EB1A for AI Researchers
EB1A requires extraordinary ability and sustained acclaim. NIW applies a different framework and may be viable even where an AI researcher does not yet satisfy EB1A.
Common AI NIW Problems
- claiming all AI work is nationally important;
- relying only on broad government statements about AI;
- failing to define the applicant’s own proposed endeavor;
- assuming an advanced STEM degree guarantees NIW eligibility;
- listing GitHub statistics without explaining technical significance;
- relying on employer prestige;
- failing to distinguish team achievements from personal contributions; and
- failing to address all three Dhanasar prongs.
Primary Legal Sources
- USCIS Policy Manual, National Interest Waivers
- Matter of Dhanasar, 26 I&N Dec. 884 (AAO 2016)
- 8 C.F.R. § 204.5(k)
Contact The Messersmith Law Firm to request an NIW evaluation.
This page provides general immigration information and is not legal advice.