Situating Legal Rationalities and Regulatory Regimes Inside AI Design and Deployment Lifecycles

This chapter reconceptualises AI development as a juridified socio-technical praxis in which legal rationalities are inscribed into data pipelines, model architectures, and deployment workflows. Integrating rule of law, human-rights, risk-society, ethics-by-design, and epistemic-justice constructs, it shows how risk-based taxonomies, supervisory regimes, and compliance-by-design infrastructures co-determine which systems are engineered and legitimised. Generative AI and foundation models are analysed as infrastructural stress tests for liability, causation, intellectual-property, privacy, and dignity, motivating alternative responsibility grammars across provider-centric, deployer-centric, and commons-oriented ecosystems. The chapter culminates in a governance toolkit that aligns public authorities, firms, technical teams, and civil society around principled design, contestability, and systemic-risk mitigation, reframing compliance as a creative modality of AI innovation.