Beyond the black-box algorithm, the Tulsa experience reshapes AI engineer training and data sovereignty

When Daphne Wang, a mechanical engineering student at MIT, joined the summer internship program within the health department of the Muscogee (Creek) Nation in Tulsa, Oklahoma, she expected her daily role to be limited to writing purely technical code. However, the practical reality required her, together with her colleague Lucy Sun, an AI and decision-making specialist, to delve into the study of tribal history and analyze its dimensions related to technological innovation, and to build the first risk-assessment monitoring system for pregnancy specific to women of the tribal nation, by dissecting data sets, reviewing literature, and designing surveys in the context of governance structures and local data sovereignty.

This experience embodies the practical pathway of the Code Tulsa program, overseen by the PKG Center for Social Impact at the institute with support from the Patrick J. McGurn and George Kaiser families. The program brings together ten students based at the University of Tulsa to provide AI and data-science solutions for local entities including the Muscogee and Cherokee Nations and community associations, aiming to merge traditional engineering mindset with a human-centric, societal perspective that understands the complexities of public challenges and their real-world contexts.

Building AI tools is no longer a technical matter isolated from data sovereignty and the privacy of local communities.This shift extends to school-focused educational initiatives, such as the Tulsa Advanced Science Camp created by Jack Carson, an electrical engineering and computer science student at the institute and a member of the Cherokee Nation, in collaboration with Ali Zhou of Harvard University. The camp attracted twenty-five students from ten tribes among a pool of one hundred sixty applicants, training them in intensive tracks covering AI engineering, DNA technologies, physics and energy, alongside philosophical workshops aimed at fostering scientific curiosity and critical thinking about the impact of technology on their communities.

This transformation in technical training curricula has direct implications for leaders of engineering schools and AI centers in the Gulf region, Egypt and the Arab Levant. The area is experiencing a rapid expansion of national genome programs, digital health-record systems and sovereign AI models. However, reliance on engineering staff whose skills are limited to model improvement and statistical validation, without a deep understanding of data-governance frameworks, community privacy and local regulations, raises the later cost of correcting systems and exposes initiatives to risks of rejection or leakage.

The most sought-after skill in the upcoming phase combines solid programming competence with the ability to engineer solutions within their cultural and regulatory environment.Involving technology students in field issues, as reflected in Tulsa’s experiences, offers a practical lesson for Arab hiring officials and academic leaders: the success of AI deployment in vital sectors such as healthcare and education does not depend solely on infrastructure quality, but on training engineers capable of making technical decisions that align with community sovereignty and actual needs.