Requirements engineering for countering hostile narratives: a vision for socio-technical intervention systems beyond detection

Abstract Hate speech narratives represent a significant societal challenge, amplified by digital platforms, algorithmic recommendation systems, and sophisticated online communication. Existing approaches to counter such narratives focus on content detection, moderation, and compliance, often treating hostile expressions as isolated artefacts rather than broader socio-technical processes. The motivation behind this vision paper is based on the premise that hate speech narratives are produced and sustained through interactions among stakeholders, institutions, technologies, and social structures. Countering hate speech is not only a linguistic or computational challenge but also a socio-technical, ethical, legal, and organisational one, since the meaning and impact of hostile expressions depend on context, speaker intentions, platform dynamics, and wider social consequences. This paper argues that countering hate speech in the broader spectrum of hostile narratives requires a holistic Requirements Engineering (RE) perspective capable of integrating technological, linguistic, ethical, legal, and societal concerns. Building on goal-oriented requirements engineering, socio-technical systems analysis, argumentation, archetypes modelling, and systems thinking, the paper outlines a requirements-driven modelling framework to monitor, analyse, mitigate, and counter hostile narratives. The framework is grounded in the definition of hate speech characteristics developed through twin lenses: the delineation of expressions and the analysis of structural and behavioural characteristics of hate speech. By positioning the countering of hate speech narratives as an RE challenge, the aim is to stimulate new research directions that bridge technical innovation with societal values and democratic governance.