Anthropic CEO Dario Amode<0xA0>Amodei has warned that the\nincreasingly rapid development of artificial intelligence (AI) poses\nsignificant risks. He expressed concern that a swarm of AI agents could\npotentially take over the entire internet within the next 6 to 12\nmonths.<\/p>\n
Amodei shared these concerns in an essay titled \u2018We Must Pace the\nFrontier\u2019, posted via X on Saturday (12\/9). He urged AI companies to\nslow the rate of AI model capability increases to allow more time to\nanticipate the risks of technology misuse.<\/p>\n
\u201cWe must slow the pace of increasing AI model capabilities. Progress\nwill still appear fast, and we must use the time we gain wisely,\u201d Amodei\nwrote in the essay, as quoted by Reuters on Thursday (17\/9\/2026).<\/p>\n
According to him, the current acceleration of AI capabilities is\nalready causing serious concern, especially since several AI models have\nproven capable of hacking external systems during various security\ntests. \u201cGiven the accelerating rate of AI capability development, my\nfear is that in 6-12 months, a swarm of such [AI] could be capable of\ntaking over the entire internet, potentially causing hundreds of\nbillions of dollars in damage,\u201d said Amodei.<\/p>\n
Anthropic previously revealed another instance where its AI models\nhacked external systems. Last July, several of the company\u2019s Claude\nmodels were reported to have breached the systems of three companies\nduring cybersecurity testing.<\/p>\n
Concerns regarding AI capabilities have also intensified following\nthe resignation of Anthropic researcher Jacob Coxon. He noted that those\nseriously involved in building AI believe the technology has the\npotential to wipe out all of humanity by the end of this decade.<\/p>\n
Amodei emphasised that his call is not to halt model training or AI\ntechnological progress. Rather, he wants companies to allow enough time\nto align and secure models before capabilities are further enhanced, and\nto involve independent evaluators to ensure safety measures are\nfunctioning.<\/p>\n
He proposed three steps, including the permanent presence of\nthird-party evaluators within leading AI companies. These evaluators\nwould be granted access to relevant internal risk assessment tools and\nprocesses.<\/p>\n
This call comes as pressure to control AI risks increases, while the\nindustry simultaneously faces strong pressure to continue technological\ndevelopment. Companies such as OpenAI and Anthropic, for instance, are\npreparing for initial public offerings (IPOs).<\/p>\n
Every new AI capability can help companies attract funding, justify\ninfrastructure investments, and strengthen IPO prospects.<\/p>\n
Amodei stated that AI companies should cooperate voluntarily to\nestablish safety standards. He believes coordination is necessary so\nthat companies can conduct safety research and build safeguards without\nlosing their competitive positions.<\/p>\n
However, he noted that slowing AI development in democratic nations\nmust still account for competition with China. Amodei stated that\nChina\u2019s AI advantage could pose a risk to United States national\nsecurity.<\/p>\n
\u201cRegulating the pace in democratic countries will be limited by the\nadvantage held by US companies over authoritarian regimes, particularly\nthe Chinese Communist Party,\u201d he wrote.<\/p>\n
Amodei also called for strengthened controls over advanced AI chips,\nmodel distillation, and the theft of model weights to prevent China from\nclosing the AI capability gap.<\/p>\n
\u201cI believe all leading AI companies must partner with governments to\nformalise the idea of permanent evaluators placed within companies, to\nprevent and document internal alignment incidents such as those\noccurring in recent months, as well as to implement regulations that\nbalance capability and safety,\u201d Amodei wrote.<\/p>", "url": "https:\/\/jawawa.id\/newsitem\/ai-boss-panics-claims-creation-could-dominate-internet-within-6-12-months-1789606409", "image": "" }, "sponsor": "Okusi Associates", "sponsor_url": "https:\/\/okusiassociates.com" }