The use of artificial intelligence (AI) in road safety depends not only on technical capabilities but also on data quality and accessibility, inter-agency coordination, and clear legal arrangements. This study identifies institutional and legal challenges to AI adoption in road safety in Iran and explores perceived relationships among them. Data were collected through semi-structured interviews with 20 experts selected purposively from four groups: traffic police, academia, knowledge-based companies working on intelligent transport systems, and knowledge-based companies active in the road sector. The reported thematic analysis produced 45 initial codes, 19 subthemes, and five main themes. This article examines the institutional and governance theme and the legal and regulatory theme. Participants described weak coordination, difficulties in data sharing, and unclear responsibilities as interconnected barriers. A review of selected Iranian legal instruments also indicates that some duties concerning the registration of crash data and access to related databases are already provided for in existing laws and regulations. The issue should therefore not be framed simply as an absence of law; rather, the adequacy, coherence, implementation, and applicability of existing rules to AI-enabled road-safety systems require assessment. The article presents an interpretive framework and policy recommendations concerning legal review, data governance, and inter-agency coordination. The proposed relationships are interpretive propositions grounded in interview accounts, not empirically tested causal relationships.