US, EU, China, and Russia differ on what AI is and how to govern it

In short: Major powers are struggling to cooperate on AI because they do not agree on what it is for, how safe it is in war, and what rules should limit it.

What's going on

Countries with the most influence over AI, including the US, the European Union, China, and Russia, often talk past each other in three areas. First, they disagree on what AI is in strategic terms. Some see it mainly as a broad tool that boosts productivity across the economy, like electricity or the internet. Others treat it more like a military advantage, or as something that needs strict international limits.

Second, they disagree on how much AI should be trusted in military settings. Several security studies argue that war still involves confusion, mistakes, and deception, so AI cannot remove uncertainty. In simple terms, even a very smart system can be fed misleading information, like a person falling for a convincing scam. That is why these sources say human judgment and safeguards still matter.

Third, they disagree on governance, meaning the rules and values that shape how AI can be used. US and EU officials have criticized AI tied to social scoring and social control, which is often seen as a reference to China’s approach to using data for public monitoring. Other research points to deeper differences over state power, surveillance, and human rights.

Some analysts also say cooperation is slowed by disagreements over which harms matter most. For example, some governments focus more on far future risks, while others focus on today’s problems like bias and discrimination.

What to watch

Experts suggest a first step could be agreeing on basic definitions and shared language, and being clearer about what each side is doing. Without that, competition and secrecy are likely to keep growing, especially in military uses.

Source: NYTimes