Anthropic has called for a slowdown in the development of artificial intelligence due to the risk of losing control — WSJ
Anthropic, the American company that developed Claude—one of the most powerful artificial intelligence systems—has called on the international community to discuss the possibility of slowing down or temporarily suspending the development of state-of-the-art AI models.
The company cites the risk of systems emerging that can independently create and improve their own versions without human intervention as the reason for this move. Anthropic believes that technological development is approaching a stage where artificial intelligence could begin recursive self-improvement, which would significantly complicate efforts to control its capabilities.
The company notes that such a scenario has not yet materialized, but the pace of the industry’s development necessitates taking potential risks seriously. According to Anthropic, as of May 2026, over 80% of the code entering the company’s production environment is generated by the Claude model. In addition, engineers’ productivity has increased eightfold over the past two years thanks to the use of artificial intelligence.
Anthropic proposes creating an international monitoring and verification system to track compliance with potential restrictions on the development of the most powerful AI models. The company compares future AI regulation to mechanisms for controlling strategic weapons.
At the same time, the initiative has sparked debate among representatives of the tech industry. Some experts believe that increased regulation could strengthen the position of the largest market players and create additional barriers for smaller companies and open-source projects.
Against the backdrop of the rapid development of artificial intelligence, the issue of its regulation is increasingly moving beyond the tech sector and becoming a matter of international policy and security.
Source: The Wall Street Journal.
As a reminder, the Russian hacking group Fancy Bear, which is linked to Russian military intelligence, hacked more than 280 email accounts of government and military institutions in NATO countries and the Balkans.
Prior to this, hackers gained access to data from the booking website Booking.com, the company reported. The attackers were able to obtain information about customer bookings, and some users received emails on Sunday notifying them of a possible data breach.
