The Iberian orca subpopulation (Orcinus orca) is Critically Endangered, with approximately 40 individuals exposed to cumulative anthropogenic pressures. Since May 2020, some individuals within this subpopulation have repeatedly interacted with sailing vessels in the western Strait of Gibraltar and Gulf of Cádiz, causing seven confirmed vessel sinkings through October 2025 but no human fatalities to date. Existing information systems including citizen-science reporting, traffic-light risk maps, official monitoring, and research campaigns remain valuable but discontinuous, reactive, and fragmented across institutional boundaries. This policy perspective argues that continuous passive acoustic monitoring (PAM), implemented through a phased edge-AI network in the Strait, could provide the missing independent detection layer needed for both orca conservation and maritime safety. Rather than treating these aims as competing priorities, the article frames PAM as shared infrastructure: A basis for behavioral monitoring, non-invasive data collection, vessel-targeted risk awareness, and a cross-jurisdictional Common Operating Picture. The required hardware and edge-AI architecture are already available, but operational warning use would require Iberian-orca-specific model validation, local propagation testing, regulatory approval, and staged field trials. The central challenge is therefore not only technical but institutional: building a governance framework that integrates marine scientists, public authorities, sailors, fishermen, whale-watch operators, and conservation stakeholders into a common system for managed coexistence. AI Use Disclosure: During preparation and editing the author used Claude (Anthropic) as a writing assistant on some portions of this manuscript. Gemini (Google) and ChatGPT (OpenAI) were used for post-completion fact and consistency checking. All intellectual content, arguments, and conclusions were developed by the author.