JupiteX Get the app
Science & Technology18 Nov 2026 · about 7 min

US-China AI race causing global ‘alarm,’ UN's top tech enovy says

The brief

The most striking warning was that countries beyond the United States and China view their AI contest with “some kind of alarm.” UN tech envoy Amandeep Singh Gill said they do not want their agency diminished while attention centers on two players. Gill said these countries want to build their own AI models and infrastructure. They are not necessarily trying to match the frontier labs in the United States and China, but still want the ability to innovate. Their concern is both about having a role in AI’s development and about how the race may affect safety rules. Gill said a race can encourage participants to ignore rules, and that recent AI safety incidents have raised concerns that “shortcuts are being taken.” The warning reflects countries’ unease about the pace and direction of AI development. Gill’s comments point to a broader challenge: pursuing innovation while keeping safety rules in view, rather than letting competition between two countries dominate the conversation.

01

What did the UN tech envoy say the rest of the world finds alarming about the US-China AI race?

The most striking warning was that countries beyond the United States and China view their AI contest with “some kind of alarm.” UN tech envoy Amandeep Singh Gill said they do not want their agency diminished while attention centers on two players.

Gill said these countries want to build their own AI models and infrastructure. They are not necessarily trying to match the frontier labs in the United States and China, but still want the ability to innovate. Their concern is both about having a role in AI’s development and about how the race may affect safety rules.

Gill said a race can encourage participants to ignore rules, and that recent AI safety incidents have raised concerns that “shortcuts are being taken.” The warning reflects countries’ unease about the pace and direction of AI development. Gill’s comments point to a broader challenge: pursuing innovation while keeping safety rules in view, rather than letting competition between two countries dominate the conversation.

02

Why can a race to develop AI lead countries to worry that safety rules will be ignored?

A race to develop AI can make speed and competition feel more urgent than following safety rules. UN envoy Amandeep Singh Gill said that “if you have a race, then you tend to ignore the rules.” The concern is not simply that countries want to innovate; it is that a contest can make safeguards seem like obstacles.

The article connects this concern to recent incidents around AI safety. Gill said these events have sparked concern among several countries that “shortcuts are being taken.” The source does not identify the incidents or describe specific rules, so it does not establish which safeguards were missed.

The issue matters because countries want to develop AI while retaining guardrails. Gill’s warning suggests that competition should not push safety aside. Countries watching the US-China contest are concerned that attention on the race could weaken respect for rules, even as other nations seek to build their own models and infrastructure.

03

What are AI safety guardrails, and what are they meant to prevent?

In general use, AI safety guardrails are rules, checks, and safeguards intended to reduce risks as AI is developed and used. They can set limits on how systems are built or deployed and help ensure that risks are considered. The article uses the term “guardrails” but does not give a formal definition or name specific measures.

Their purpose is to prevent unsafe practices and reduce the chance that AI systems cause harm. In the article, Gill links guardrails to the danger of a race: when competition puts pressure on participants, they may ignore rules or take shortcuts. The source does not list particular harms or describe a specific guardrail.

Gill said recent AI safety incidents have prompted concern among several countries that shortcuts are being taken. His comments make guardrails important not only as technical protections, but also as rules that should remain part of AI development during competition. This broader definition draws on general knowledge; the article itself focuses on countries’ concerns about the race.

04

Why do countries want to build their own AI models and infrastructure, even if they are not competing with the biggest labs?

Countries want their own AI models and infrastructure because they do not want their agency taken away while global attention focuses on the United States and China. Gill said that does not mean they are trying to compete with those countries’ frontier labs. They still want to innovate and build capacity of their own.

The article describes this as a desire to develop models and infrastructure within their own countries. That aspiration goes beyond joining a race to produce the most advanced systems. For countries such as Ethiopia, Gill said, building capacity also involves access to GPUs, connectivity, and energy.

Gill’s comments show that the AI contest has consequences beyond the two leading players. Other countries want access to emerging technologies and the ability to develop their own approaches. The article does not specify what models they plan to build or how their infrastructure will be organized. It does emphasize that building capacity is important for bridging the AI divide and enabling broader innovation.

05

Besides GPUs, what does a country need to build AI capacity?

GPUs are not the only requirement for building AI capacity. Gill said countries also need connectivity and energy availability. His point is that having computing hardware alone does not create the capacity needed to build and use AI; other supporting resources must be available too.

The article gives Ethiopia as an example of a country that needs a “massive infusion of capacity.” Gill said it needs significantly more GPUs, but immediately added that capacity alone is not enough. Connectivity and energy are also necessary. The article does not provide figures for the amount of electricity, network access, or computing power needed.

This matters to Gill’s broader concern about bridging Africa’s AI divide and guaranteeing the continent access to emerging technologies. Building infrastructure therefore means addressing more than hardware. The article identifies connectivity and energy as requirements, but does not specify projects, costs, or a timeline. It presents investment in African AI as a way to create a market and support useful applications.

06

What is the scale of Ethiopia’s AI-capacity challenge, according to the article?

Ethiopia’s scale is central to the capacity challenge Gill described: its population is above 130 million. He said a country of that size needs significantly more GPUs. The article does not state how many GPUs Ethiopia currently has or how many it would need, so it gives no precise hardware target.

Gill described the need as a “massive infusion of capacity.” But he stressed that adding computing capacity alone would not be enough. Connectivity and energy availability are necessary too. The example underlines that building AI infrastructure involves several requirements, not just acquiring GPUs.

Gill raised Ethiopia while emphasizing the broader need to address Africa’s AI shortfall. He said more capacity needs to be built to bridge the AI divide and guarantee the continent access to emerging technologies. The article provides no budget or timetable for meeting that need. It also suggests investment could create a market and support applications in areas such as water scarcity, public health, and public services.

07

What is a GPU, and why do AI systems need so much computing power?

A GPU, or graphics processing unit, is a kind of processor that can handle many calculations at once. This parallel processing makes GPUs useful for the heavy computing involved in training and running many AI systems. The article does not define GPUs; this explanation draws on general knowledge.

AI systems need computing power to process data and perform the calculations involved in their development and use. Using GPUs can provide some of that capacity, which helps explain Gill’s emphasis on Ethiopia needing significantly more of them. The article does not give a specific GPU count or describe which AI systems Ethiopia plans to build.

Gill also stressed that GPUs alone will not solve the capacity challenge. Connectivity and energy availability are necessary as well. His example sits within a wider call to build Africa’s AI infrastructure, bridge the AI divide, and ensure access to emerging technologies. The article’s point is that computing hardware is one important part of capacity, not the whole foundation.

This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.

Read more in the JupiteX app

Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.

Or read more news on the web