Play store
The rapid advancement of artificial intelligence (AI) is outpacing governments’ ability to regulate the technology, leaving countries increasingly vulnerable without reliable scientific evidence to guide policymaking, according to a newly released report by the United Nations’ Independent International Scientific Panel on Artificial Intelligence.
The panel officially launched its Preliminary Report on July 1, offering the first global, independent scientific assessment of AI’s opportunities, risks and societal impacts. The report is intended to provide governments with an evidence-based foundation as they prepare for the inaugural UN Global Dialogue on AI Governance in Geneva on July 6-7 and ahead of the panel’s comprehensive assessment, scheduled for publication in 2027.
The report comes at a time when AI technologies are evolving at an unprecedented pace, transforming economies, public services, education, healthcare and communication. While these developments promise significant benefits, they also pose growing challenges for governments struggling to keep regulations aligned with rapidly changing technology.
According to the panel, policymakers around the world are making critical decisions amid uncertainty, relying on rapidly evolving and sometimes conflicting sources of information. The report argues that sound scientific evidence is essential to developing effective AI governance frameworks that reflect national priorities while addressing global concerns.
Prepared by 40 experts from diverse disciplines and every region of the world, the report represents a broad scientific effort to establish a shared understanding of AI. The panel includes computer scientists, economists, legal scholars, academics, human rights specialists and other experts who serve independently of governments, companies and institutions.
The report’s findings will be presented to government representatives during the first UN Global Dialogue on AI Governance, where policymakers are expected to discuss international cooperation and future approaches to regulating AI.
Speaking during a virtual briefing following the report’s release, panel co-chair Yoshua Bengio said AI is entering a phase where its rapidly expanding capabilities could have profound consequences for humanity.
“Intelligence gives power,” Bengio said. “As that power grows, it can unlock tremendous benefits if managed responsibly, but it can also create significant risks if left without adequate safeguards.”
Bengio stressed that the panel’s role is to provide objective scientific evidence rather than prescribe specific policies. Maintaining independence from political debates, he said, allows the panel to present credible assessments that governments can use in developing their own regulatory approaches.
Co-chair Maria Ressa, the journalist and Nobel Peace Prize laureate, said the panel’s diverse membership often brought different perspectives to the discussion, but members found common ground through scientific analysis.
She noted that AI-driven technologies have contributed to increasingly fragmented information environments, making it more difficult for societies to distinguish between fact and misinformation. The report, she said, aims to establish a shared factual foundation upon which governments can build effective AI governance.
One of the report’s central findings is that AI capabilities have accelerated dramatically in recent years, alongside widespread adoption across industries and public institutions. However, governments have struggled to keep pace with these developments, creating regulatory gaps that could expose societies to growing risks.
The report emphasizes that scientific evidence should form the basis of AI governance. Without reliable evidence, policymakers may struggle to understand the technology’s implications or respond effectively to emerging challenges.
Despite these concerns, the panel highlights AI’s considerable potential to advance sustainable development. AI applications are already improving healthcare, education, agriculture and food production in many parts of the world. When adapted to local conditions and supported by appropriate training, AI systems can enhance productivity, improve public services and expand access to essential resources.
The report notes that countries in the Global South have demonstrated promising uses of AI in sectors such as health and agriculture. However, successful implementation depends on tailoring technologies to local needs, institutional capacities and cultural contexts rather than simply adopting solutions developed elsewhere.
At the same time, significant inequalities remain in access to AI technologies. According to the report, many developing countries continue to lag behind advanced economies in AI adoption, investment and governance.
The panel found that 118 countries, most of them in the Global South, are not actively participating in major international AI governance discussions. In addition, fewer than one-third of developing countries have established national AI strategies.
These disparities raise concerns that developing nations could become increasingly dependent on technologies designed elsewhere while having limited influence over international standards governing AI development.
The report also identifies major imbalances in global AI infrastructure. Of the 500 largest known public and private AI computing clusters worldwide, approximately 75 percent are located in the United States, 15 percent in China and the remaining 10 percent across the rest of the world.
Similarly, AI model development is concentrated among a relatively small number of companies. Around 91 percent of notable AI models originate from the private sector, highlighting the growing influence of large technology firms in shaping the future direction of AI.
The concentration of computing resources and technological expertise, the report warns, could reinforce existing global inequalities by allowing only a handful of countries and corporations to shape AI standards, economic opportunities and technological innovation.
Maria Ressa warned that decisions affecting humanity’s technological future are increasingly being made by a limited number of powerful actors.
The report also raises concerns about AI-generated content and its impact on public trust. Advances in generative AI have made it easier than ever to produce convincing text, images, audio and video, making it increasingly difficult for people to distinguish authentic content from manipulated or synthetic material.
Such developments could facilitate the spread of misinformation and disinformation, undermine confidence in public institutions and weaken democratic processes. The report also notes growing evidence that AI systems may disproportionately affect minority communities if they are trained on biased or incomplete data.
Looking ahead, the panel outlines multiple possible paths for AI development. Continued rapid progress could deliver significant economic and scientific benefits but also amplify social risks if governance fails to keep pace. Alternatively, slower technological progress could provide governments with additional time to strengthen regulatory frameworks and build institutional capacity.
The report concludes that international cooperation will be essential to ensuring AI serves the public interest. Governments are encouraged to strengthen scientific expertise, invest in national AI strategies and participate more actively in global discussions on AI governance.
As world leaders prepare to gather in Geneva for the first UN Global Dialogue on AI Governance, the report provides an evidence-based foundation for discussions on how countries can balance innovation with safety, fairness and human rights.
UN Secretary-General António Guterres urged governments not to delay action, warning that the faster AI advances without shared international rules, the less influence governments and citizens will have over its future direction. He said the scientific evidence is now available, and the responsibility lies with policymakers to translate that knowledge into effective governance that benefits all of humanity.