Musk Sees Irony in Anthropic Potentially Creating the AI Superintelligence It Warns Against: ‘The Most Entertaining Outcome Is the Most Likely’

Elon Musk highlighted the irony of Anthropic potentially developing the very artificial superintelligence it fears, responding to an online warning about the company’s AI safety ambitions.

Anthropic Warns About Risks Of Superintelligence

In an X post on Monday, Musk wrote, "The most entertaining outcome is the most likely, especially if ironic," He was responding to a post imagining Anthropic "building the super-intelligence that they fear," thereby validating its own warnings. Musk did not explicitly predict that outcome.

The most entertaining outcome is the most likely, especially if ironic https://t.co/CJD71WTJQq

— Elon Musk (@elonmusk) November 10, 2026

The exchange follows growing concerns about AI systems improving themselves faster than researchers can evaluate their behavior. In its ‘Responsible Scaling Policy’ update in mid-August, Anthropic assessed catastrophic risks from increasingly capable models and described its safeguards.

Its Oct. 19 safety roadmap then warned that automated AI research could accelerate development faster than governance can keep pace. The company is preparing proposals addressing development speed and independent evaluations.

Reuters reported last week that research nonprofit METR found Anthropic’s models improving approximately 1.9 times faster than before as AI tools assisted researchers. The organization stressed that its estimate remained uncertain.

Anthropic Is Also Musk’s Major Commercial Customer

The exchange also highlights an unusual business relationship. In May, Anthropic agreed to use Space Exploration Technologies Corp.’s (NASDAQ:SPCX) Colossus 1 computing facility, gaining 300 megawatts of capacity from more than 220,000 Nvidia processors.

SpaceX, in an SEC filing later that month, disclosed a computing contract requiring Anthropic to pay $1.25 billion monthly through May 2029, subject to cancellation provisions. Anthropic has since become SpaceX’s largest commercial customer, accounting for reportedly 34.5% of third-quarter revenue.

The arrangement makes Musk both a potential vocal critic of Anthropic and a major supplier of computing capacity powering its Claude models.

AI Safety Debate Shapes Washington’s Regulatory Agenda

A separate Reuters report last Friday noted that Democrats plan to introduce legislation requiring independent safety evaluations of advanced AI models and clarifying companies’ liability when their technology causes harm.

Meanwhile, Anthropic’s Advanced AI Framework, a policy proposal that details mandatory government safety standards, compute thresholds, and societal resilience investments, calls for mandatory safety testing, independent evaluations, and government authority to block dangerous AI deployments. The company also opposes broad federal measures that would override stronger state-level AI protections.

Image via Shutterstock