African emergency care systems: AI-aware but not AI-ready Authors: Willem Stassen Published: 2026 Whether we are ready or not, Artificial Intelligence (AI) has entered our emergency care system. It is here to stay, and scaling. The survey by Tadesse and colleagues, published in this issue of the AfJEM, provides one of the first glimpses of how facility-based emergency physicians and trainees experience this reality [1]. In my view, one thing is clear: We are AI-aware, but we are not AI-ready. Results of the survey indicate that most respondents report having a basic understanding of AI (88%), are aware of AI tools and their use in emergency care (73%), and many had used AI tools in their practice before (73%) [1]. An important limitation of this study, however, is that 68% of respondents were from Ethiopia – the survey therefore provides us with a small slice, and cannot represent the picture for the whole continent. Notwithstanding, of concern is that the infrastructure to support responsible and efficient use of AI (or readiness) is almost entirely absent: only 14% of respondents had formal training in AI use, and only 12% of respondents had workplace strategies in place to support AI use and implementation [1]. While uptake and understanding seem good, AI’s actual utility is a lot more modest than what these numbers report because the majority of AI use in the workplace seems to be non-clinical for research and writing, versus truly clinical for things like AI triage and clinical analytics. It is the latter AI applications that will transform emergency care on the continent. That means we still have time to influence AI development on our own terms. What ought AI readiness and AI development look like for African emergency care? Respondents of the survey rightly cite ethical concerns with its use and call for more training and more policy, [1] but these interventions are wholly insufficient. True readiness requires key factors that are largely absent on the continent: validated African datasets to train AI models; [2] functioning and reliable digital infrastructure; African-led governance frameworks; and tools designed for our diverse African contexts rather than imported into them. Lastly, tools should be affordable, and openly licensed so that new dependencies on commercial entities head-quarted in high-income countries are not created. AI tools that are developed within commercial, profit-driven ecosystems will make access prohibitively expensive [3]. Without these interventions, AI will accelerate, not solve inequity, and become a sophisticated solution to problems that were mis-defined by others and validated on data from settings that bear no resemblance to our realities. In this way, perhaps, the more important question to ask is not whether African clinicians are ready for AI, but rather, whether the AI deployed is ready for Africa. An AI tool that is ready for Africa is one built on data we provide, to solve problems we define, and governed according to our values. The current AI governance landscape reflects a kind of epistemic injustice, in which African values and knowledge systems are absent from the frameworks that will ultimately determine how AI behaves on the continent [4]. The values that non-African AI governance frameworks are built upon tend to value individual autonomy and algorithmic transparency. However, on the continent, our values are rooted in communal identity and relational accountability. The Ubuntu ethic, which conceives of personhood as fundamentally relational, offers a philosophical foundation for AI governance that is distinctly African and deeply relevant to emergency care [5]. These principles ought to shape how triage algorithms are evaluated, how patient data are governed, and how the clinician–patient relationship is protected from automated decision-making. For example, an AI-based trigger tool would not only focus on individual diagnostic accuracies but would also consider the communal impact of these decisions, and call for collective accountability. Ethics, after all, is a human invention. It cannot be outsourced. It cannot be imported. Seventy-eight percent of respondents on this survey expected that AI will be used routinely in emergency departments in the future – and I think they are right. AI has arrived in our emergency care system, but on whose terms, governed by whose values, and in service of whose patients? If the answer is not a resounding “our own,” then African emergency care clinicians might be ready for AI, but the AI is not ready for Africa. Declaration of interest statement The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.