Research / Sources
Legal research on AI governance. law-ai.org
Authors assess evidence that automating AI R&D could trigger an intelligence explosion, analyse its risks including loss of control and eroded checks on power, and propose policy responses.
Argues that frontier labs should train many AIs under diverse constitutions rather than one value set, citing risk mitigation, legitimacy, emergent value and avoiding value lock-in.
Argues that frontier AI use inside the executive branch poses structural risks to the rule of law, such as weaker resistance to unlawful orders and evasion of review, and sets out a research agenda.
Christoph Winter's remarks to the European Parliament on risks AI agents pose to democracy, recommending law-following AI architectures and a better-resourced EU AI Office.
Argues that highly capable AI could cause catastrophic harms in Global South countries through weak institutions and other vulnerabilities, and that stakeholders must act now to influence its development.
Argues that the Global AI Majority is largely excluded from AI development and governance, and proposes interest alignment, participatory architecture and safety assurance to address dual-use and frontier-state restrictions.
Argues that liability is a central but limited tool for AI governance: it addresses risk externalities but not public-good problems like alignment research funding or diffuse structural harms, which need subsidies and other policies.