▲ Anthropic Claude
As artificial intelligence (AI) technology has advanced rapidly in recent years, a widening policy vacuum is emerging among governments worldwide, the U.S. daily The New York Times (NYT) analyzed on September 27 (local time).
According to the NYT, the rapid evolution of AI technology has left regulatory and oversight capabilities of governments in the United States, China, the European Union (EU), and almost everywhere else struggling to keep pace, making it difficult to control the technology.
The NYT pointed out that the EU's AI regulatory act passed two years ago has limited effectiveness, while many other AI regulatory bills around the world remain stalled without coming to a vote.
Although response measures have been discussed at international summits, they are already being quickly forgotten.
The NYT noted that AI has widened the gap between policymaking and technology to an unprecedented level, stating, "Many governments are caught in conflict, wanting to harness AI while fearing its risks."
It added that the situation has grown more complex due to domestic politics, limited technical expertise, geopolitical divisions, and fears of falling behind in a race that could determine economic power and national security for decades to come.
Many nations are watching the policies of the United States and China—the two leading AI superpowers—to determine their direction, but both countries themselves lack certainty.
Within the U.S. government, voices emphasizing the need for regulation for AI safety are at odds with concerns over diminishing industrial competitiveness.
In particular, U.S. President Donald Trump has expressed negative views on regulations, arguing they stifle innovation and warning of the risk of ceding leadership to China due to such rules.
China has introduced some AI regulations centered on chatbot content censorship, but it is simultaneously pursuing development at a rapid pace.
Cooperation between the United States and China, the two leading AI powers, is crucial for establishing international safety standards, yet both nations are competing to prevent the other from gaining a technological edge.
During the summit between President Trump and Chinese President Xi Jinping held last week, the two leaders failed to agree on a safety protocol.
Furthermore, at the UN General Assembly in New York last week, European Commission President Ursula von der Leyen and leaders from over 20 countries signed a letter urging joint action for AI safety, but President Trump stated that the U.S. will reject "any attempt to build globalist plans" to control AI technology.
In the U.S. federal legislature, bills have been introduced to establish a federal agency to oversee AI models, mandate kill switches, and ban the most powerful models exceeding human capabilities, but active pushes for votes have been lacking.
Amid these circumstances, policy vacuums are worsening across nations, and AI companies such as Anthropic and OpenAI are largely left without external safety regulatory frameworks.
The EU passed its AI regulation, the Artificial Intelligence Act, in 2024, but years have passed with the implementation of key provisions put on hold, and critics point out that the law needs revision to reflect technological advancements even before it can be properly implemented.
The initial draft of the EU's AI Act emerged in early 2021, 18 months before OpenAI released ChatGPT.
The draft AI Act focused on issues of the time, such as facial recognition software, and was subsequently revised to align with the rise of general-purpose AI systems like ChatGPT.
European Parliament member Valérie Hayer stated in a recent press conference regarding the AI Act, "When we adopted this law, we could not even imagine what AI can do today."
She stated that while the law was ahead of its time, "we must adapt."
French President Emmanuel Macron pointed out last year that Europe has been left out of AI technology development, saying, "If we let the U.S. and China monopolize all the leading companies, one thing is certain: We will have the best regulations in the world, but we will have nothing left to regulate."
India, Japan, Australia, Canada, and Kenya have opted for more relaxed approaches or wait-and-see stances, while Brazil is discussing more comprehensive regulations.
(Photo: AP, Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.