by 
Key Takeaways
- The EB-2 NIW lets AI and deep tech founders self-petition for a U.S. green card without an employer sponsor or PERM labor certification, but the standard tightened after USCIS’s January 2025 Policy Manual update.
- Qualifying turns on the three-prong Matter of Dhanasar test: substantial merit and national importance, being well-positioned to advance the endeavor, and showing the U.S. benefits on balance from waiving the labor certification.
- AI, advanced computing, and semiconductors sit squarely on the federal Critical and Emerging Technologies (CET) list, which USCIS treats as a strong positive factor under the Policy Manual (Vol 6, Part F, Ch 5).
- Premium processing is available for $2,965 (effective March 1, 2026) and requires USCIS action within 45 business days.
- RFE and denial rates have risen. In FY2023, 90% of STEM EB-2 denials were NIW cases, so evidence framing matters as much as credentials.
The EB-2 National Interest Waiver (NIW) has become the most discussed permanent residence pathway for AI and deep tech founders in 2026. The reasons are practical: no employer sponsor required, no PERM labor certification, premium processing on a 45-business-day clock, and a USCIS policy framework that explicitly favors work in critical and emerging technologies.
The reasons it has also become the most misunderstood pathway are equally practical. Petition volume tripled after the January 2022 STEM policy update, and USCIS responded with sharper scrutiny. A January 2025 Policy Manual revision raised the evidentiary bar for the “well-positioned to advance” prong, and adjudicators are now issuing detailed Requests for Evidence on the same fact patterns that sailed through three years ago.
For an AI or deep tech founder, the question is no longer whether the field qualifies. It does. The question is whether the petition is constructed in a way that survives current adjudication standards.
The Legal Framework: Dhanasar and the Three-Prong Test
Every EB-2 NIW petition is adjudicated against the framework set out in Matter of Dhanasar, 26 I&N Dec. 884 (AAO 2016). A petitioner must establish each of the following:
- The proposed endeavor has substantial merit and national importance.
- The petitioner is well-positioned to advance the proposed endeavor.
- On balance, the U.S. would benefit from waiving the job offer and labor certification requirements.
The Dhanasar test is the same for a postdoctoral researcher in materials science and a Y Combinator-backed AI founder. What differs is the evidence each can put forward, and how that evidence maps to the prongs.
Prong One: Substantial Merit and National Importance
Substantial merit is rarely contested for AI work. National importance is where most petitions are won or lost. USCIS distinguishes between work that benefits a single company or set of clients and work whose impact reaches a national field, industry, or strategic priority.
For deep tech and AI founders, the federal Critical and Emerging Technologies (CET) list, maintained by the National Science and Technology Council, is the most useful objective anchor. The current list includes:
- Artificial Intelligence (including foundation models and machine learning)
- Advanced Computing
- Semiconductors and Microelectronics
- Autonomous Systems and Robotics
- Quantum Information Technologies
- Advanced Communications and Networking
- Cybersecurity
USCIS Policy Manual Volume 6, Part F, Chapter 5 directs adjudicators to treat work furthering a CET field as a positive factor, particularly when the work bears on U.S. economic competitiveness or national security.
The CHIPS and Science Act of 2022 provides additional statutory support for founders in semiconductor, advanced manufacturing, and AI hardware niches.
Executive Order 14179 (January 2025), which directs federal agencies to remove barriers to American leadership in AI, has been cited in petitions to underscore current administration priorities, though it is not a substitute for case-specific evidence.
Prong Two: Well-Positioned to Advance
This is where the January 2025 Policy Manual update raised the bar. USCIS now expects a clear, evidence-backed showing that the petitioner can move the endeavor forward. For founders, that typically means:
- Capitalization evidence (priced rounds, SAFEs, grant awards)
- Accelerator or incubator admission (Y Combinator, Techstars, NSF SBIR, AI2 Incubator)
- Issued or pending patents and peer-reviewed publications
- Pilot deployments, design partners, or government contracts
- Independent expert letters tied to specific contributions, not generic praise
A common failure pattern is a strong founder with a thin written business plan. The Policy Manual treats the business plan as central evidence for entrepreneurs, and an underdeveloped plan invites an RFE even when the founder’s credentials are otherwise strong.
Prong Three: Balance of Factors
The third prong asks whether the U.S. is better served by waiving labor certification for this specific person than by requiring the standard process. For founders, the argument is structural: a founder cannot realistically sponsor themselves through PERM, and requiring a labor market test would undermine the very entrepreneurship the U.S. has identified as a national priority.
What Has Changed for 2026 Petitioners
Three shifts are worth flagging:
- Higher RFE rates. Most RFEs cite the second prong, asking for more specific evidence of execution capacity. Founders who file early, perhaps before product traction or funding, should expect this.
- Closer scrutiny of the “proposed endeavor” definition. USCIS expects a concrete, narrowly framed endeavor, not a broad description of the founder’s field. Rather than a broad statement like ‘building AI infrastructure,’ an ideal proposed endeavor is specific and measurable. For example: ‘Developing and deploying a domain-specific foundation model for clinical decision support, with measurable adoption across U.S. health systems.
- Premium processing remains 45 business days at a $2,965 filing fee, but it does not affect the visa bulletin. For Indian and Chinese nationals, the rate-limiting step remains country-of-chargeability backlogs, not the I-140.
How Alcorn Immigration Law Approaches These Petitions
Alcorn Immigration Law is based in Silicon Valley and represents AI and deep tech founders across the EB-1A, EB-2 NIW, O-1A, and related pathways. The firm’s Legal Launch program is a structured four-step process: an initial evaluation of the founder’s evidentiary profile against multiple categories, a milestone-based plan to strengthen the weakest prong, case architecture and filing, and longer-term immigration positioning across status changes and dependents.
The work is less about checklist-driven document collection than about constructing a coherent legal narrative, one in which the proposed endeavor, the evidence of advancement, and the national-interest argument reinforce each other across hundreds of pages of exhibits.
The EB-2 NIW remains one of the most direct routes to U.S. permanent residence for AI and deep tech founders in 2026, but “direct” is not “easy.” Approval increasingly depends on how carefully the petition is constructed against an adjudication standard that has tightened since 2022. To discuss whether an EB-2 NIW fits your situation, contact Alcorn Immigration Law to schedule a consultation.
Frequently Asked Questions
Can an AI founder qualify for an EB-2 NIW without a Ph.D.?
Yes. A U.S. master’s degree or foreign equivalent in a relevant STEM field satisfies the advanced-degree professional threshold. A bachelor’s degree plus five years of progressively responsible experience can also qualify. The exceptional-ability route, which requires meeting at least three of six regulatory criteria, is a separate pathway and may suit founders without traditional degrees but with substantial professional recognition.
Does my AI startup need to be incorporated and funded before I file?
No, but funding and incorporation can serve as strong evidence on the second prong. USCIS evaluates whether you are well-positioned to advance the endeavor, and capitalization, accelerator admission, customer pilots, or government grants substantially strengthen that showing. Filing earlier is possible, but petitions filed without traction usually require stronger compensating evidence — patents, publications, or detailed expert opinion letters tying you to specific innovations.
How long does the EB-2 NIW process take in 2026?
Standard I-140 adjudication ranges from approximately 8 to 25 months, depending on service center workload. Premium processing reduces the initial action window to 45 business days for a $2,965 fee. After I-140 approval, the timeline to a physical green card depends on visa bulletin availability, which is current for most countries but heavily backlogged for India and China.
Is the EB-2 NIW better than the O-1A visa for founders?
They serve different purposes. The O-1A is a nonimmigrant work visa with a lower evidentiary threshold and faster processing, but it does not lead to permanent residence on its own. The EB-2 NIW is an immigrant petition leading directly to a green card. Many founders use both — O-1A first to establish a U.S. presence, then EB-2 NIW or EB-1A to pursue permanent residence.
What counts as “national importance” for an AI founder?
National importance means the endeavor’s prospective impact extends beyond a single employer or set of clients. Building proprietary internal tools for one company is unlikely to qualify on its own. Developing a foundation model with cross-industry applications, improving U.S. cybersecurity infrastructure, or advancing semiconductor design tied to CHIPS Act priorities will generally satisfy this prong if properly documented.
Can I self-petition while on an H-1B or O-1?
Yes. The EB-2 NIW is a self-petition, meaning you file Form I-140 on your own behalf regardless of your current employer-sponsored status. Maintaining valid nonimmigrant status during the process is important for those who plan to adjust status inside the U.S. via Form I-485 rather than pursue consular processing abroad.
What are the most common reasons EB-2 NIW petitions are denied?
The most frequent denial grounds are vague proposed endeavors, insufficient evidence of execution capacity (the second prong), and generic expert letters that praise the petitioner without tying them to specific contributions. Petitions that read like résumés rather than legal arguments are particularly vulnerable.
Does the 2025 USCIS policy update make it harder for founders to qualify?
The January 2025 update raised the evidentiary expectations on the second prong, particularly around the credibility of business plans and the specificity of the proposed endeavor. It did not change the underlying Dhanasar framework. Founders with substantive traction and well-constructed petitions are still being approved at rates consistent with the broader STEM NIW population.
This article is intended for general informational purposes only and does not constitute legal advice. Immigration law is fact-specific, and you should consult a qualified attorney about your individual circumstances.
