UK House of Lords proposes AI kill‑switch powers

A coalition of investors and AI researchers has announced a new safety initiative that it says will act as a “vital safety net” against runaway artificial‑intelligence systems from companies such as OpenAI and Anthropic. The project, dubbed the AI Resilience Initiative, is backed by a mix of technology firms, philanthropic foundations, and university research groups that collectively pledge more than $200 million in initial funding. Its stated goal is to develop independent oversight mechanisms, safety protocols, and audit teams that can evaluate and mitigate risks posed by large language models and other advanced AI systems.

The initiative will focus on three core areas: alignment research to ensure AI behaves in ways that match human values, robustness testing to guard against unexpected failures, and governance frameworks that enable transparent accountability. Grants will be awarded to academic labs and independent think tanks that work on these topics, while a central oversight board will coordinate the evaluation of new AI deployments. The backers argue that existing corporate safety measures are insufficient on their own, and that an external, neutral safety net is needed to prevent potential misuse or unintended consequences of rapidly evolving AI technologies.

By positioning itself as a complementary layer to corporate and governmental safeguards, the AI Resilience Initiative seeks to influence industry standards and policy discussions. While its long‑term impact remains to be seen, the launch signals a growing recognition among AI stakeholders that proactive, independent safety mechanisms are essential as the field moves toward increasingly powerful and autonomous systems.