Big tech, the state, or you? Who is responsible for generative AI addiction?

The rise of generative AI introduces significant social and ethical challenges, notably the potential for addiction, mirroring problematic social media and gaming use. Preliminary evidence suggests harmful addictive properties that may lead to individual, social, and societal harms. This article addresses the research questions: Who or what can be held responsible for generative AI addiction, and what consequences can arise from the resulting responsibilities? We first conceptualise generative AI addiction and review initial empirical evidence for problematic use, then employ the conceptual lens of responsibility, reviewing precedents from other addiction areas (substance and behavioural) to inform the governance of generative AI. The paper proposes a model for a responsible AI ecosystem where accountability is distributed across four key stakeholder groups: governmental bodies, tech companies, researchers, and civil society. We argue that tech companies, who stand to gain most, hold a key role and must engage proactively to mitigate risks, cautioning that reluctance to accept these responsibilities may incur severe reputational and financial costs, similar to those faced by the tobacco and pharmaceutical industries. This work contributes to the technology addiction and AI governance discourses by proposing collaborative accountability structures for emerging technologies.