Dario Amodei, the CEO of Anthropic, released a statement urging the imposition of speed limits on the advancement of advanced AI technologies. He proposed a three-phase strategy with a focus on 'verifiability': Initially, companies ought to engage third-party evaluators for on-site assessments. Subsequently, they should foster coordinated efforts between the United States and its allies. Finally, they should aim for worldwide cooperation. He highlighted that AI has already started to demonstrate recursive self-improvement capabilities, and within the next 6 to 12 months, swarms of intelligent agents could potentially gain control over the internet via botnets. Hence, it is crucial to decelerate the enhancement of model capabilities and utilize the additional time to bolster operational reliability, alignment, interpretability, testing, evaluation, and other safety-related aspects. He opined that the United States must sustain and widen its technological edge over China to create room for deceleration, and that global speed restrictions will ultimately necessitate China's involvement. Additionally, he suggested a four-tier international agreement framework, although he conceded that a broad-based speed limit is unlikely to be implemented in the near future. OpenAI's CEO, Altman, reposted the statement and voiced his support, while Musk also concurred. Earlier, Jacob Cockx, a researcher at Anthropic, cautioned upon his resignation that AI might pose an existential threat to humanity within the next decade.