A preliminary report from the United Nations Independent International Scientific Panel on Artificial Intelligence warns that unchecked AI development could lead to severe harms, including “catastrophic” outcomes. Multiple outlets report that the panel views the current pace of AI capabilities as outstripping the ability of governments and regulators to respond effectively. The panel says policymakers face a dilemma: they need robust scientific evidence to design appropriate safeguards, but evidence and assessment methods may not be keeping up with rapid technological change. The report also highlights safety concerns related to AI system behavior and use. These include risks from more autonomous systems that could be harder to control, as well as deception risks such as “lying” or deceptive outputs, which can be exploited by malicious actors. Several sources emphasize that the report is a preliminary assessment and frames the findings as an early, independent scientific input ahead of broader international discussions. Overall, the sources agree the panel is urging attention to risk management and regulation informed by evolving scientific understanding.