The UN's Independent International Scientific Panel on Artificial Intelligence released its preliminary report, warning that AI development is advancing too quickly for current governance systems to manage effectively. The panel found that AI capabilities double in complexity every few months, creating a regulatory nightmare for policymakers worldwide.

While AI offers tremendous benefits like accelerating drug discovery and enabling early disease detection, it also enables serious harms. The report documents cases of AI-generated deepfakes, including child sexual abuse material, and notes how these systems can spread misinformation or reinforce dangerous user behaviors. California investigated Grok over nonconsensual deepfakes, highlighting real-world consequences.

Policymakers face a fundamental challenge: regulations typically require scientific data, but AI systems evolve faster than researchers can study them. This leaves authorities constantly playing catch-up with technology that's already moved on to the next iteration.

The panel recommends stronger independent evaluation, international cooperation, and common standards to keep AI safe and accountable. It notes that access remains heavily concentrated in the US and China, leaving developing nations behind. The UN will host the Global Dialogue on AI Governance in Geneva starting July 6 to address these gaps.