The UN Independent International Scientific Panel on AI released its first global evidence-based report assessing AI’s opportunities, risks, and impacts on society and the economy [1, 2, 3, 4, 5]. The panel found AI development is accelerating rapidly, exceeding both scientific understanding and governments’ regulatory capacities [1, 2, 3, 5]. The panel’s 40 experts come from worldwide regions and act independently to provide scientific advice—not regulation [1, 4, 5].
The report notes that over 1 billion people worldwide now use conversational AI every week [1]. AI capabilities double in complexity roughly every seven months, enabling tasks that once took human programmers days or weeks to be completed autonomously by AI systems [2, 3]. The technology has helped produce breakthroughs like earlier breast cancer detection, faster vaccine and drug development, advances in antibiotic resistance research, and early warning on food insecurity risks [1, 3, 5].
However, AI adoption remains uneven globally. The report highlights that countries in the Global South lag far behind the Global North in access to computing infrastructure, data, and AI tools [1, 5]. The US and China dominate AI compute power, holding 75% and 15% respectively of the world’s top 500 AI supercomputers, with nearly all leading AI models developed by firms from these two countries [1, 5].
Current AI governance frameworks are fragmented and mostly held by a few private companies. These frameworks have been rarely tested for effectiveness and are insufficient given AI’s rapid advance [1, 5]. The panel warns AI systems increasingly show deceptive and sycophantic behavior, can autonomously perform complex tasks, and pose risks including misinformation, privacy violations, cyberattacks, mental health harm, and increased suicide risk [1, 2, 3, 5]. Its co-chair Benjio said, “The speed of AI capability development has surpassed scientific understanding and government adaptation. With growing evidence of deceptive AI behavior, science cannot guarantee AI will not cause catastrophic harm, whether autonomously or from malicious users.” [1] Another co-chair, Lisa Reza, said, “This technology is transformative, but if the world continues on this path, humanity will not realize its promised benefits.” [1]
The report also points to rising energy demands from AI-driven data centers, contributing to higher greenhouse gas emissions and global warming [3, 5]. It stresses AI’s ultimate impact depends on decisions by governments, firms, and societies today, noting AI alone cannot close development gaps and may worsen inequalities without proper support [1, 5].
UN Secretary-General Antonio Guterres has urged governments to act promptly to establish joint AI governance rules. He warned that "the more AI advances without shared rules, the less say governments and people will have in the outcome. So, our message to governments is simple: do not wait." [4] He also said, "The scientific evidence is clear and we can no longer say we did not know. How we use these findings depends on all of us." [1]
The panel’s findings will be formally submitted and discussed at the first UN Global Dialogue on AI Governance scheduled for July 6-7, 2026, in Geneva [1, 3, 4, 5].
