Former OpenAI and Anthropic Researcher Warns AI Race Could Lead to Extinction by 2030

Jacob Coxon, a former AI researcher at both OpenAI and Anthropic, warns that the current pace of artificial intelligence development could lead to human extinction by 2030. Coxon resigned from Anthropic after concluding that the industry is prioritizing speed over survival.

Coxon, who previously worked on pretraining models at OpenAI, alleges that the world's leading AI labs are engaged in a reckless race toward "self-improving superintelligence" without implementing adequate safety measures. "Neither company is acting responsibly," Coxon stated. "They are racing to self-improving superintelligence and gambling with our lives."

The Push for a Slowdown

This alarm comes as internal pressure mounts within "frontier" AI labs regarding the safety of autonomous systems. These concerns are not limited to former employees; Anthropic CEO Dario Amodei has also publicly called for a slowdown in development. Amodei described the growth of AI as "exponential" and noted that the trajectory is "starting to get steep," though he clarified that this does not mean there is a need to "panic today."

These warnings are part of a broader movement to establish industry-wide safety standards and external evaluations. A significant coalition of industry insiders has already begun organizing; over 1,100 employees from OpenAI, Anthropic, Google, and Meta signed a letter known as the "Pacing the Frontier" initiative. The group is calling for the implementation of government-led mechanisms to slow the development of AI to ensure safety keeps pace with capability.

Why the Warning Matters

Coxon's resignation carries particular weight because he has operated within the inner workings of the two most prominent AI labs in existence. His public alarm suggests a growing gap between the public-facing "safety culture" of these corporations and the actual technical trajectory of their models.

If the race to superintelligence continues unchecked, critics argue it could create a "race to the bottom" where safety is sacrificed for competitive advantage. The involvement of high-level researchers suggests that the risk of catastrophic outcomes is no longer a theoretical exercise for philosophers, but a concrete concern for the engineers building the systems.

The Path Forward

As the industry moves toward models capable of autonomous self-improvement, the focus now shifts to whether governments can effectively regulate a technology that evolves faster than legislation. The primary remaining question is whether the "Pacing the Frontier" initiative can secure the necessary political will to enforce a slowdown, or if the competitive pressure between labs will override the warnings of their own creators.