JMIR Publications

ABSTRACT

As Ai companies and governments race toward innovation, regulation is essential. Scientists and academicians, as stewards of knowledge and leaders in technological innovation, must also develop the ethical compass and resource consciousness needed to guide Ai use. Generative Ai is becoming part of everyday research practice, while shared norms for appropriate use remain unsettled. We argue that health services researchers should not be passive consumers of this transition but should help define the scientific and professional culture surrounding Ai. Ai literacy, human agency, and superuse provide a means for users to govern how Ai is used and harness its benefits responsibly. Ai literacy provides a foundation for informed user agency and responsible Ai use. Responsible use must include consciousness of the environmental resources consumed through Ai use. We introduce Ai superuse as proactively building the skills and strategies to use Ai proficiently, deliberately, and responsibly to maximize benefit to ourselves and others. Further, Ai is an implementation—and de-implementation—challenge. Earlier virtual health implementation efforts focused on usability and promoting adoption. Generative Ai presents a different challenge: adoption is widespread and increasingly embedded in routine research tools. Health services research should apply its implementation expertise not only to practices worth adopting and scaling, but to de-implement uses that are unreliable, inefficient, inequitable, unethical, or unnecessarily resource-intensive. Existing guidance converges around principles including transparency, safety, equity, and human accountability, while standards for disclosure and use remain inconsistent. We call for shared disclosure standards, usable guidance, human oversight, environmental stewardship, and communities of practice that bring governance into everyday Ai use and give users practical means to exercise agency and harness its benefits responsibly.