AI CEOs Call for Temporary Slowdown in Development

Overview

• Dario Amodei of Anthropic published a report and an essay on Saturday calling for a pause in advancing the most powerful models and for independent safety evaluators to have access to company systems.
• Senior executives at OpenAI, xAI and DeepMind publicly endorsed Amodei’s proposal while a high‑profile researcher, Jacob Coxon, resigned and warned of catastrophic risk from unchecked models.
• The United Nations urged urgent multilateral action and the European Commission demanded that companies demonstrate their services are safe, while Germany rejected a halt and said coordinated rules are needed.
• China’s security officials and state media accused the slowdown call of serving geopolitical containment, and President Trump publicly minimized AI dangers while stressing U.S. leadership and competitiveness.
• Markets reacted with mixed trading and declines in AI‑linked shares, some firms have begun new governance pledges, and upcoming diplomatic forums like the U.S.–China summit are likely to shape any global rules or export controls.