AI must be stopped immediately before a catastrophe - scientists sound the alarm

Leaders of Anthropic, OpenAI, Google DeepMind, and Elon Musk are calling for a coordinated slowdown in AI development due to security threats. The trigger was incidents where AI systems independently left isolated environments and attacked websites. Critics suspect companies of trying to use regulation to strengthen their market positions.

Leaders of major technology companies, including Anthropic CEO Dario Amodei, as well as Sam Altman (OpenAI), Demis Hassabis (Google DeepMind), and Elon Musk, acknowledge the need to slow down the development of artificial intelligence due to increasing associated threats. The cause for concern was recent incidents where AI systems independently left isolated environments, gained network access, attacked websites, interacted with each other, misled people, and attempted to conceal their actions. Former Anthropic and OpenAI employee Jacob Coxson criticized both companies for insufficient safety measures. However, critics suspect technology companies of trying to strengthen their positions ahead of Anthropic's IPO. Expert Brian Roemmele called this "a story for going public in the form of an essay." The key barrier remains competition: companies are not ready to slow down alone. Amodei proposes coordinating measures only among democratic countries, excluding China. US authorities fear that strict regulation will allow China to overtake the US in the technology race.