- On February 24, 2026, on Nikhil Kamath’s “WTF Is” podcast, Anthropic CEO Dario Amodei revealed that Claude 1 existed in 2022, before ChatGPT’s launch.
- However, the company decided not to release it immediately, fearing it would trigger an “AI arms race.”
- Amodei argued that releasing a powerful model too early could lead competitors to rush their products without adequate safety testing.
- He called the decision to hold back Claude 1 a “one-time overhang”—a moment where Anthropic had the advantage but chose not to exploit it.
- Claude 1 was subsequently officially released in March 2023.
- The consumer AI competition only truly exploded when OpenAI launched ChatGPT in late 2022.
- Amodei said the delay gave “the world a few more months” to prepare before the race accelerated.
- However, the commercial price was immense: Anthropic may have lost its leading position in the consumer AI market.
- He admitted this decision caused the company to “cede” its initial advantage to rivals.
- Beyond model release issues, Anthropic has also spoken out on chip policies and AI regulations, even if it displeased some suppliers.
- The company also expressed views differing from the U.S. government on certain AI policies.
Conclusion: Dario Amodei stated that Claude 1 existed before ChatGPT in 2022, but Anthropic deliberately delayed its release to avoid sparking an uncontrolled AI race. Although this may have cost them their commercial advantage and leadership in consumer AI, the company believes it gave the world more time to address safety issues. Anthropic is also willing to challenge both chip suppliers and government policies to maintain its stance.
