AI4SE and SE4AI: A decade review of AI in systems engineering

H. Sinan Bank, Daniel R. Herber and Thomas Bradley submitted a 10-page paper titled "AI4SE and SE4AI Exploration: A Decade Looking Back and Forward" to arXiv on 17 Jun 2026. The authors trace progress in artificial intelligence and systems engineering across three phases and publish a human-AI agreement dataset plus a web explorer to let readers compare relevance judgments.

What does the paper map and what does it find?

The paper describes three phases the authors label foundational, applied, and LLM inflection, and it identifies five critical research gaps while offering guidance for practitioners. The March 2020 INCOSE INSIGHT special issue on AI and Systems Engineering is cited as a pivotal moment: it became the most downloaded issue in that publication's history and launched a research community that now draws over 250 registrants to its annual workshop. The authors report their conclusions across those three phases and point readers to the AI4SE/SE4AI Explorer web application and shared agreement data for deeper inspection.

How was the literature review performed and what scale was covered?

The authors combined human expertise with six AI models to perform a human-AI agreement literature review assessing 1,712 INCOSE INSIGHT articles and 889 SERC publications. That review produced relevance judgments and agreement data the paper shares alongside the web application so readers can compare their own relevance decisions with the human and AI raters. The methods and scale are presented in the paper's 10 pages and illustrated across five figures.

What are the paper's concrete outputs?

Concrete outputs enumerated in the submission include the agreement dataset and the AI4SE/SE4AI Explorer web application. The manuscript itself spans 10 pages and includes five figures. The authors supply an arXiv DOI and the paper is classified under Artificial Intelligence, Digital Libraries, and Systems and Control. The combination of mapped phases, the literature-scale review, and the released tools form the paper's primary deliverables.

Why it matters

The paper ties a visible community milestone, the March 2020 INCOSE INSIGHT special issue, to measurable community growth: an annual workshop with over 250 registrants. By bringing together human experts and six AI models to assess large corpora (1,712 and 889 items), the authors move beyond anecdote toward reproducible evidence about what literature is relevant to AI-for-systems-engineering and where consensus breaks down. Practitioners facing adoption, assurance, and workforce questions gain both synthesized gaps and a tool to test their own judgments against the paper's raters.

What to watch

Watch for updates to the AI4SE/SE4AI Explorer and any follow-up studies that reuse the shared agreement data. The submission date on arXiv is 17 Jun 2026; subsequent tooling updates, workshop programs tied to the INCOSE community, or published follow-ups that address the paper's five identified gaps will be the clearest signals that the field is acting on this work.