EB-1A for AI and Software Engineers

The Criteria Are Field-Neutral

The ten evidentiary categories in 8 CFR 204.5(h)(3) are written to span "the sciences, arts, education, business, or athletics." They speak of awards, original contributions "of major significance," scholarly articles, judging the work of others, high remuneration, and a leading or critical role — not particularly of professorships or academia. An engineer who has never written an academic paper can still satisfy several of them.

The regulation goes a step further. Under 8 CFR 204.5(h)(4), when a listed criterion "does not readily apply" to your occupation, you may submit comparable evidence to establish eligibility. USCIS has also emphasized, in its policy guidance, examples of how professionals in science and technology can meet these standards. For industry engineers, comparable evidence is often the difference between a criterion that looks academic on its face and one you can actually document.

What "Major Significance" Actually Asks

The most valuable — and most misunderstood — criterion for engineers is original contributions of major significance. A high GitHub star count, a granted patent, or a fast-growing library is a starting point, not a finish line. AAO decisions consistently look past the existence of a contribution to its documented impact on the field: adoption by others, citations or references in a professional context, licensing, or implementation beyond your own employer. Expert letters help, but on their own they are treated as opinion; the persuasive records pair them with objective corroboration.

The same discipline applies across the board. A senior title does not by itself prove a leading or critical role — the record has to show the role mattered relative to others and that the organization has a distinguished reputation. High pay is strongest when it is compared to a credible benchmark for your field and location, not asserted in isolation. For more on this evidentiary bar, see our deeper dive on original contributions of major significance and on expert letters USCIS finds persuasive.

Meeting Three Criteria Is Step One

EB-1A is adjudicated in two steps (the framework from Kazarian v. USCIS). First, USCIS counts how many of the ten criteria your evidence satisfies; you need at least three. Then it steps back for a final merits determination — a totality-of-the-record judgment about whether you have sustained national or international acclaim and are among the small percentage who have risen to the top of the field. Clearing three criteria is necessary, but it does not by itself win the case. Engineers who assemble a tidy checklist and stop there are the ones who draw a Request For Evidence on the merits. Our article on why meeting three criteria isn't enough walks through this second step in detail.

Landing an EB-1A is not out of reach for industry. It simply means building the case around how your evidence reads together — the same way it would for a STEM researcher or any other industry professional. Preparation is the difference between a record that happens to be strong and one that is presented to be strong.