IMF and Anthropic Call for Shared Responsibility in AI Governance

The International Monetary Fund and artificial intelligence company Anthropic have called for governments, businesses and communities to share responsibility for managing risks created by rapidly developing AI systems.

Speaking at an IMF-World Bank Annual Meetings panel, IMF Managing Director Kristalina Georgieva said AI governance requires broader participation than technology companies alone. She urged major AI-producing countries, including the United States, China and India, and other participants in AI supply chains to help establish safeguards. Georgieva compared the need for common safety standards to protections used in Formula One racing.

Anthropic co-founder and head of public benefit Jack Clark said companies should not make governance decisions by themselves. He said governments are establishing AI security institutes and conducting tests, while Anthropic permits third parties to evaluate its frontier models and publishes risk assessments. Clark also cited emerging agreements against using AI for destructive activities such as biological weapons or cyberattacks.

Both emphasized the need for AI companies to share data that can help governments respond to labor-market effects. Georgieva said IMF research indicates demand for skills in new or transformed higher-paying jobs, while repetitive tasks are more likely to disappear. Clark said AI may have little effect on employment in some sectors, making detailed data important for policymaking.

The calls come amid wider debate over AI oversight, including how companies should respond when chatbot conversations indicate possible violence without creating broad surveillance of users.