World leaders discuss AI safety amid rising concerns and US-China summit

President Donald Trump and Chinese President Xi Jinping are set to meet in Washington for a three-day summit that will include discussions on artificial-intelligence (AI) risks. On the same day, CEOs of leading AI companies, including OpenAI chief Sam Altman, urged the United Nations Security Council to establish safeguards, warning that uncontrolled AI could threaten humanity.

Governments and industry groups have already created a patchwork of safety measures: national AI safety institutes test new systems before deployment, voluntary corporate pledges aim to protect elections and curb deepfakes, and the G7-led reporting mechanism allows AI firms to share model details and develop risk-reduction standards. An annual scientific report also tracks emerging threats.

Experts say the existing framework lacks a real-time, global response capability. Proposals under discussion include a 90-day, opt-in rapid-response protocol that would trigger a confidential incident report within 72 hours, a standardized pre-release disclosure system modeled on the G7 Hiroshima AI Process, and a US-China AI-incident hotline that could later be expanded to other nations.

While the suggested steps draw on established mechanisms such as the OECD’s AI Incidents Monitor and post-crisis stress-test models, critics note they may favor Western perspectives and rely heavily on existing institutions. Nonetheless, analysts argue that immediate, practical actions are needed to address accelerating AI development and growing geopolitical tensions.