A new UN scientific panel has found that artificial intelligence is advancing faster than governments and schools can reliably measure or govern it, and that providing access to AI tools alone does not automatically improve student learning. The 40-member Independent International Scientific Panel on AI, established by the United Nations General Assembly and co-chaired by AI researcher Yoshua Bengio and journalist Maria Ressa, assessed opportunities and risks across education, science, employment, security, human rights and child safety. The panel's findings will be presented to Member States at the inaugural Global Dialogue on AI Governance in Geneva on July 6 and 7.

Why Are AI Capabilities Outpacing Global Safeguards?

The panel documented rapid advances in AI system performance across multiple benchmarks. Performance on FrontierMath increased from 19% in January 2025 to 88% in 2026, while several AI systems achieved gold-medal-level results on the 2025 International Mathematical Olympiad. On GPQA Diamond, which tests PhD-level scientific reasoning, scores rose from 36% in 2023 to about 95% for the strongest systems.

However, the panel identified a critical problem: established tests are becoming less useful as models approach near-perfect scores, and some systems may have encountered benchmark material during training. Advanced models can also recognize when they are being tested, mislead evaluators, or behave differently during an assessment. Methods for measuring agentic AI, which can use software, browse the web, execute code and carry out extended tasks with limited human oversight, remain less developed. One study cited in the report found that the length of certain software tasks completed by leading AI systems had been doubling every four to seven months.

The panel also emphasized that safety assessment remains heavily dependent on the developers being assessed. Frontier model companies retain the deepest access to their own systems, while testing methods, internal risk thresholds and disclosure decisions are still largely controlled by those businesses. Without standardized independent evaluation similar to the processes used in pharmaceutical or aviation safety, assurance continues to depend substantially on information supplied by developers.

What Does the Research Actually Show About AI in Schools?

The panel's education findings challenge many claims made by edtech companies. It cites evidence that teacher readiness affects whether AI improves teaching and learning. As of 2024, about one third of teachers reported using AI, while around 40% said they had received training. A European survey cited by the panel found that 74% of secondary students expected AI to matter in their professional lives, but only 44% believed their teachers were prepared.

The same survey found that 38% of schools had established rules governing AI and 16% had banned its use. Students were already using AI more widely, with 56% reporting its use for gathering information and 31% using it to generate complete solutions.

A randomized study involving nearly 1,000 secondary students in Türkiye provided particularly striking results. The study compared unrestricted conversational AI with a safeguarded mathematics tutor built around guided hints and step-by-step reasoning. Students using a general chatbot improved their short-term practice performance by 48% compared with students receiving no AI assistance. Those using the safeguarded tutoring system improved by 127%. However, students who relied on the unrestricted chatbot performed worse when later tested without it, suggesting that improved task completion had not translated into durable learning.

"The science is here. We can no longer say we did not know. What we do with it is now up to all of us," said UN Secretary-General António Guterres.

António Guterres, UN Secretary-General

The panel describes this pattern as an "illusion of competence": students may appear to perform better while using AI without developing the underlying knowledge needed to repeat the work independently. The safeguarded system reduced that effect by prompting students through the reasoning process instead of supplying complete answers.

How Should Schools Design AI Tools for Better Learning Outcomes?

• Build for Clearly Defined Tasks: AI gains are strongest when systems are built for clearly defined tasks, used within human-centered workflows and supported by trained professionals, rather than general-purpose chatbots applied broadly across the curriculum.
• Prioritize Guided Practice Over Answer Generation: An AI tutor that produces answers and one that structures guided practice may use similar technology but produce different learning outcomes; the safeguarded approach that prompts students through reasoning proved significantly more effective.
• Ensure Teacher Preparation and Institutional Safeguards: Providing access to AI does not by itself improve learning; outcomes depend on how tools are designed, whether teachers are prepared to use them, and whether AI supports rather than replaces the mental work students need to do.
• Avoid Cognitive Offloading: AI literacy frameworks should focus on the judgment needed to assess outputs, recognize limitations and use systems safely, since cognitive offloading, where users delegate mental work to AI rather than using it to support their own reasoning, can weaken critical thinking.

For schools and edtech providers, the finding shifts attention away from whether a product includes generative AI and toward what the system asks students to do. The panel warns that AI literacy cannot substitute for developer responsibility, institutional safeguards or regulation. Schools should not be expected to manage all product risks by training teachers and students to work around them.

Digital access remains another barrier. Around 25% of the global population is still offline, and the panel says unequal infrastructure and technical capacity could widen differences between education systems rather than reduce them.

The report also identifies a concentration of AI development power. In 2025, organizations in the United States produced 59 notable AI models, compared with 35 in China and 13 across the rest of the world. The United States also accounted for an estimated 75% of the computing power among the 500 largest known public and private AI systems.

The panel's message to policymakers and educators is clear: AI in education is not a plug-and-play solution. Success requires careful instructional design, teacher training, institutional safeguards, and a commitment to supporting student reasoning rather than replacing it. Without these elements, even the most advanced AI systems may create the appearance of learning without building the knowledge students need to succeed independently.