July 12, 2026, (Inside AI) — A landmark United Nations scientific panel, co-led by Turing Award winner Yoshua Bengio and Nobel laureate Maria Ressa, has issued a stark warning: artificial intelligence is evolving so rapidly that governments risk losing the ability to govern it. The preliminary report, released by the UN Secretary-General, marks the first comprehensive global scientific assessment of AI and sets the stage for a more detailed analysis next year.
The 40-member Independent International Scientific Panel on AI found that current safety measures and governance frameworks are failing to keep pace with breakthroughs in frontier models and autonomous agents. These systems promise productivity leaps but introduce risks that existing institutions cannot yet manage.
“The world cannot govern what it cannot understand,” UN Secretary-General António Guterres said at the report’s launch, emphasizing the need for independent scientific evidence to guide policy. The report examines AI across seven themes: science advances, applications in healthcare and education, economic impacts, security, human rights, cultural wellbeing, and governance.
Policymakers face a painful dilemma: wait for scientific certainty and risk severe harm, or act on incomplete knowledge. The panel argues that waiting is no longer an option, as harms are already materializing in areas like disinformation and labor displacement.
The Compute Chasm: How Infrastructure Shapes AI Sovereignty
The global AI race is no longer just about talent—it’s about who controls the computing hardware. The report reveals a staggering concentration: the United States holds roughly 75% of global AI computing capacity, China 15%, and the rest of the world just 10%. This compute divide turns processing power into a strategic resource, not merely a technical input.
In 2025, U.S. institutions produced 59 notable AI models, China 35, and the rest of the world only 13. The same year, 75% of the computing power in the 500 largest AI clusters sat in the U.S., 15% in China, and 10% elsewhere. Such dominance risks locking nations without advanced chips and data centers into permanent consumer status, unable to build sovereign AI or shape global standards.
For countries like India, which has launched the IndiaAI Mission to procure compute hardware, the findings underscore the urgency of domestic infrastructure. Without it, the panel warns, digital inequalities could harden into long-term technological dependence, leaving nations vulnerable to decisions made in Silicon Valley or Beijing.
Monopoly by Design: The Risks of Corporate and Geopolitical Capture
Beyond hardware, the AI value chain—from semiconductor design to cloud platforms—is consolidating in a few firms and nations. The report states: “The concentration of AI capabilities in a small number of firms and countries could enable authoritarian capture and undermine democratic accountability.” It adds that current systems reflect only a narrow slice of linguistic and cultural diversity, excluding much of the global population.
This concentration extends to evaluation and oversight. The panel notes that dozens of governance instruments exist but are fragmented, concentrated among a few corporations, and rarely measure real-world effectiveness. “Evaluation methods themselves are underdeveloped, and the institutions needed to provide independent capability and risk assessments remain embryonic,” the report says.
The Global South faces disproportionate risks from AI misuse, with limited local resilience. The panel calls for proactive investment in public-interest AI, open research, and wider access to compute and scientific resources. Without broader participation, AI’s benefits and decision-making power could reinforce existing inequalities, leaving governments with little strategic autonomy in critical sectors like defense and healthcare.
The preliminary report is the first in a series of periodic assessments, with a more comprehensive version expected next year. It serves as both a diagnostic and a call to action, urging governments to build independent evaluation capacity now—before the next generation of AI systems makes current governance debates obsolete.
