Abstract Public debate about AI shifts between existential alarm and complacent reassurance. We read this volatility through the classic lens of moral panic—a discourse that amplifies diffuse fears and pulls policy toward symbolic gestures—and set it against a pragmatic program that treats risks as diagnosable and governable. Reconstructing AI’s shift from task-specific tools to an inference infrastructure clarifies why anxiety spikes: authority has moved upstream to data curators, model/compute providers, and benchmark communities, while outputs grow more fluent and more opaque. Rather than adjudicating the panic, we translate it into action. This paper operationalizes pragmatism by defining the actionable pathway: measurable controls, independent oversight, and cross-sector coordination that keep AI’s promised gains while tightening accountability where it matters. Its portability is demonstrated across employment, education, and ethics. Then drawing on a Neo-Triple Helix perspective, we propose a governance framework that coordinates universities, industry, government, and civil society through participation mechanisms, assurance guarantees, and procurement levers. This approach aims to translate principles into verifiable practice, grounding AI governance in evidence, inclusivity, and transparent accountability to guide AI development toward broad societal benefit.
Borkowska et al. (Mon,) studied this question.