AI Startup Giant Warns of Runaway Development, Unusually Calls for Global Slowdown on Advanced AI Research | WPN World People News

Artificial intelligence (AI) startup Anthropic has recently issued a warning that the human role in AI development is rapidly diminishing. According to internal data released by the company, more than 80% of the code in its software repositories is now generated automatically by its flagship model, Claude, while the amount of code submitted by engineers each quarter has surged to several times previous levels. Researchers are concerned that this feedback loop of technological self-development could trigger recursive self-improvement within the next two years, enabling AI systems to evolve autonomously without human intervention and potentially escape human control. To prevent severe social disruption, the company is calling on both the United States and China, as well as AI developers worldwide, to temporarily halt the advancement of the most powerful AI systems.

However, this bold proposal for restrictions has sparked intense controversy across Silicon Valley and Washington. As Anthropic is reportedly approaching a trillion-dollar valuation and pursuing a confidential public offering, many White House officials and industry competitors have questioned whether the initiative is, in fact, a marketing strategy aimed at weakening rivals under the banner of AI safety. Some senior policymakers have argued that slowing down unilaterally would effectively hand China a decisive strategic advantage in the global race for critical technologies.

Faced with a regulatory challenge comparable to nuclear weapons oversight, the absence of a reliable verification mechanism remains the central obstacle. Although U.S. President Donald Trump has signed an executive order requiring the most advanced AI models to undergo a 30-day review before release and has sought preliminary discussions with China on AI safety cooperation, the highly concealed nature of software development creates strong incentives for companies and nations to secretly continue advancing their systems. In response, Anthropic is planning to work with experts, policymakers, and other stakeholders to develop a practical multilateral framework designed to prevent violations and ensure responsible AI development.