The Pulse
UN AI panel says safeguards cannot wait for certainty
The United Nations’ first independent scientific panel on artificial intelligence warned that AI is advancing faster than public oversight, while Secretary-General António Guterres called for common testing standards and child-safety protec
AI.info Team ·
This article is based on an automatic-speech-recognition transcript published by the United Nations. The UN says the transcript is not an official record or official document.
UN panel presents its first report on artificial intelligence
The United Nations’ Independent International Scientific Panel on Artificial Intelligence presented its first report on July 6 during the opening Global Dialogue on AI Governance in Geneva.
The panel brings together 40 experts from different regions and disciplines, serving in their personal capacities and independently of governments, companies and other institutions. Its report provides an evidence-based assessment of artificial intelligence’s current effects, opportunities and risks. The programme says the document is the panel’s first report and that it does not make specific policy recommendations.
UN Secretary-General António Guterres said the report identifies three broad warnings: the speed of AI development, the concentration of computing power, data and talent, and the effect of synthetic or machine-generated information on public trust.
He said AI systems are no longer limited to waiting for direct instructions. They are writing code, acting online and making choices with less human oversight, while institutions remain designed largely to govern machines that follow commands.
Guterres also said the most advanced systems are concentrated in a small number of companies and countries, leaving many nations with little influence over decisions that will affect their futures. He warned that machine-enabled falsehoods can persuade as effectively as truthful information, while authentic evidence can be dismissed as fabricated.
Guterres calls for common testing methods
Guterres called for countries to align on methods for testing AI systems, measuring risks and assigning responsibility across borders. He said the world needs common baselines for advanced systems and common methods to evaluate and verify risks.
“No child should be a guinea pig for unregulated AI.”
António Guterres, Secretary-General of the United Nations
He also announced an AI Child Safety Pledge with three requested safeguards. Companies should not deploy systems accessible to children without child-specific safety testing and independent oversight. They should have zero tolerance for the generation of sexual images of children and must detect, report and remove such material. Systems should also connect children showing signs of distress with human support rather than handling a crisis alone.
Guterres said human responsibility must remain in place when AI is used in high-stakes decisions involving justice, health care and policing. Machines may inform those decisions, he said, but humans must decide and answer for the outcome.
Capacity remains uneven
The secretary-general said the benefits and risks of AI are distributed unevenly. Computing infrastructure, data and technical expertise are concentrated, while many developing countries lack the resources needed to participate fully in AI development or govern systems on their own terms.
He said the United Nations is working with member states on a global network for exchanging knowledge and building AI capacity, particularly in developing countries. He also said he would submit recommendations for a Global Fund for AI to support skills, data and affordable computing power.
The panel is not a regulator and cannot impose binding rules. Its role is to provide independent scientific analysis to governments and other decision-makers. The final responsibility for policy choices therefore remains with member states.
Guterres said the panel’s evidence should support urgent action rather than delay. The central question, he told delegates, is whether governments will shape AI together or allow the technology to shape societies without a shared plan.