The public has long grown accustomed to Elon Musk's bold proclamations, but this time, he is making a point that he has been right all along. The CEO of SpaceX and Tesla has voiced his endorsement of a lengthy essay published by Anthropic's CEO, Dario Amodei. Amodei's central argument is that artificial intelligence with recursive self-improvement capabilities will advance at a pace that outruns human comprehension and control, and even if such technology is to be developed, it must be approached with the utmost caution. To mitigate future risks, he proposes that frontier AI companies should host third-party assessors on-site to inspect AI models, monitor labs' adherence to safety protocols and commitments, and report safety incidents—evaluating not just the risks of finished models, but also scrutinising the alignment safety of the entire training process and technical pipelines.
On January 22, 2026, Elon Musk attended the World Economic Forum annual meeting in Davos, Switzerland, where he responded on X, stating: "Dario is right." He also noted that he has been sounding the alarm on AI risks for a long time, attaching a post from 2023. He wrote at the time: "In my view, the risk of artificial general intelligence (AGI) is far greater than that of nuclear weapons. No matter how smart a person is, it is difficult to imagine an existence that far surpasses their own intelligence."
Interestingly, while Musk issues these warnings, he is also pouring substantial funds into the AI sector. SpaceX's AI division is aggressively building data centres and rolling out the Grok chatbot, while Tesla is pushing forward with its "physical AI" projects, including autonomous robotaxis and the humanoid robot Optimus, which is slated to begin production by the end of this year. The largest current investment is concentrated in the data centre business. SpaceX's Chief Financial Officer, Bret Johnsen, recently revealed that as of December last year, the company's AI compute business had reached an annual recurring revenue (ARR) of $45.7 billion, and he is even more confident about hitting the $100 billion ARR target by the end of this year.
Musk is not the only one who would suffer revenue losses from a slowdown in AI investment. Amodei, as well as his rival OpenAI CEO Sam Altman, may both have to scale back operations and postpone listing plans. On February 19, 2026, Anthropic CEO Dario Amodei spoke at the AI Impact Summit in New Delhi, India. Anthropic's IPO filing document, the S-1, could be submitted at any time, and the document may outline a slowdown in product iteration and infrastructure expansion, which could drag on profitability metrics. However, the Financial Times reports that after excluding certain costs, the company's profit margin could still hold at 80%. Altman has already announced that OpenAI will delay its listing until 2027, citing the need to address AI safety concerns as the public reason.
The collective move by tech giants to pump the brakes on AI development raises a question: is it driven by genuine concern for humanity's long-term safety, or, as Michael Burry has questioned, a utilitarian move to inflate the valuation of their own AI assets? One thing remains certain: Musk's bets tend to pay off, which benefits his listed companies and steadily propels him towards becoming the world's first trillionaire.