S-risks are suffering risks (the term most commonly expands this way in the source literature): the possibility that future technologies, such as advanced artificial intelligence or space colonization, could create suffering on scales vastly exceeding anything in Earth's history, futures argued to be worse than extinction because they contain astronomical amounts of misery rather than mere absence. The concept comes from effective-altruism and longtermist philosophy.
In debate, s-risks appear as impact framing on technology topics: teams argue that some outcomes deserve categorical priority because suffering at that scale dominates every conventional impact. Engaging them means pressing the speculative link chains (these scenarios stack many uncertain steps) and the ethical premises (whether preventing vast suffering outweighs preserving existence is a genuine philosophical debate, not a settled assumption). As AI topics recur across events, this corner of longtermist vocabulary keeps migrating into rounds, so knowing the term is increasingly practical rather than exotic.