The rapid adoption of artificial intelligence brings not only technological progress but also a serious threat of increasing global inequality. This conclusion was reached by the UN Independent International Scientific Panel on AI, which brings together 40 independent experts. According to analysts, the key problem lies not in access to the technology itself, but in control over it.

Countries that rely on imported models and cloud infrastructure risk losing sovereignty over safety standards and the adaptation of technologies to local needs. In other words, they receive a tool, but not the power over it. This creates a situation where the benefits of AI are distributed extremely unevenly.

Computing Hegemony: 90% of Capacity in the Hands of Two Countries

The numbers speak for themselves: the United States accounts for about 75% of the computing power of the world's leading AI supercomputers, while China controls approximately 15%. Together, these two countries own nearly 90% of all computing resources used to train the most powerful systems. Such concentration, experts warn, could lead to increased authoritarian control and weakened democratic accountability. Lagging countries urgently need to develop their own infrastructure: data centers, local developers, and technical specialists.

Language inequality is a particular cause for concern. Generative models demonstrate high-quality performance with English and other widely spoken languages, while for other languages, especially rare ones, the quality of generation remains unacceptably low. The example of machine translation into the Tigrinya language is a striking confirmation: the system confused smallpox with syphilis, and gonococcus with diabetes. Such errors are not just inaccuracies — they can be deadly.

"In the long term, the gap between rapidly growing functional capabilities and the availability of effective risk management methods could lead to catastrophic results," the report notes.

We should not forget about the basic digital divide. According to the International Telecommunication Union, in 2025, 2.2 billion people remain offline. For these regions, the implementation of AI runs up against basic access to the internet, electricity, and digital skills. The report is not a regulatory document, but it offers clear guidelines for policymakers: developing local infrastructure, increasing AI literacy, investing in developers, creating safety institutions, and strategies to combat disinformation.

As an analyst, I see in this report not just a warning, but a clear signal for the crypto industry and decentralized technologies. It is blockchain and distributed computing that could become an alternative to the current monopoly on computing power. Decentralized GPU networks and protocols for training models are not just a trend, but a potential tool for democratizing access to AI, allowing countries and communities to maintain control over their data and technologies.