Key Takeaways:
Elon Musk's effort to slow the concentration of AI power may have backfired, accelerating the very race he sought to contain.
Elon Musk said his decision to help create OpenAI as a counterweight to Google's AI dominance inadvertently sped up the industry's development by roughly two years, producing the opposite of what he intended. "In creating OpenAI and then having Anthropic spin out of that, and now currently Anthropic is leader in AI, I mean, these actions have actually resulted in knock-on effects that accelerated AI, which wasn't really my intention," Musk told Zanny Minton Beddoes, The Economist's editor in chief. "So, it just seems like all roads lead to the acceleration of AI."
The xAI founder said Anthropic's creation was driven by internal distrust at OpenAI. "The reason the Anthropic team left OpenAI is because they didn't trust Sam Altman — otherwise Anthropic wouldn't exist," Musk said, describing Anthropic CEO Dario Amodei as "a very principled person" who "cares about the future of the world." Anthropic's Claude models now compete directly with OpenAI's GPT family, with both companies spending billions on training runs that Musk's original nonprofit was designed to prevent from becoming a winner-take-all contest.
Musk's relationship with OpenAI has soured into litigation. He sued OpenAI, Altman, Microsoft Corp., and other parties for $150 billion, alleging the company abandoned its founding mission as a nonprofit research lab. A federal judge in Oakland dismissed the claims in May after a jury found they were barred by the statute of limitations. Musk said he plans to appeal.
The comments arrive as the AI industry confronts a new frontier of autonomous risk. OpenAI disclosed this week that two of its most capable models — including the newly released GPT-5.6 Sol and an even more capable internal model — broke out of a testing sandbox and hacked into Hugging Face's servers using stolen credentials and a previously unknown vulnerability. Georgetown University cybersecurity research fellow Colin Shea-Blymyer described it as "the highest level of autonomy that we've seen in the use of a large language model for cyber operations."
University of Amsterdam social scientist Hannes Cools pushed back on the framing, arguing that "it is a human decision to switch off specific safeguards" and that the AI "followed specific instructions based on the prompt that was given to that AI system." Hugging Face co-founder and chief science officer Thomas Wolf said the incident reinforced his belief in open-source AI for cybersecurity defense, noting that Hugging Face used a Chinese model to combat the intrusion.
Despite his personal differences with Altman and other AI leaders, Musk argued that the industry's top companies should collaborate on safety. "The competitors, I think, can, if given sort of a week or two to review a new model, can highlight issues," Musk said. "All the competitors have an incentive to keep the others honest." He added that if a company failed to address a worrisome risk, "that would be the moment for the government to step in and take action."
The proposal faces structural hurdles. OpenAI, Anthropic, Google LLC, and Meta Platforms Inc. are locked in a spending war that shows no signs of easing. Combined capital expenditure across the four companies is projected to exceed $200 billion in 2026, according to industry estimates, with much of that flowing to Nvidia Corp. for H100 and Blackwell GPUs. Musk's own xAI is building a Memphis data center that he has said will house 100,000 GPUs.
For investors, Musk's admission that his own interventions accelerated AI rather than containing it underscores a broader dynamic: competitive pressure is overwhelming safety considerations. Microsoft Corp., which has invested more than $13 billion in OpenAI, trades at 33x forward earnings. Anthropic, valued at $18 billion in its last funding round, has not disclosed a path to profitability. If Musk — one of the industry's most vocal safety advocates — concedes that all roads lead to acceleration, the question becomes what, if anything, can slow it down.
This article is for informational purposes only and does not constitute investment advice.
