One of the founders of AI company Anthropic is warning that letting artificial intelligence develop without guardrails could trigger economic instability. The co-founder argued that regulators need to install something akin to a 'brake pedal' for the technology — a mechanism to slow or stop certain AI deployments before they cause widespread harm.
The Warning From Anthropic
Speaking about the pace of progress in the field, the co-founder described a scenario where rapid AI deployment outpaces society's ability to adapt. Jobs, financial markets, and entire industries could face sudden disruption, they said. The concern is not hypothetical — the pace of development over the past two years has already forced governments and companies to scramble for oversight frameworks.
The warning comes from inside a company that is itself pushing the boundaries of what AI can do. Anthropic develops large language models and competes with firms like OpenAI and Google. That insider perspective gives the call for regulation a particular weight, even if the company has its own commercial interests in the space.
A 'Brake Pedal' for AI Development
The co-founder used the metaphor of a brake pedal to describe the kind of regulatory tool they think is needed. Not a kill switch, they explained, but a way to ease off the accelerator when the risks become clear. The idea is that regulators would have the authority to pause or restrict certain AI systems if they pose a clear threat to economic stability or public safety.
That concept is not new in technology regulation. Countries have used similar mechanisms for pharmaceuticals, financial products, and even self-driving cars. But applying it to AI raises questions about who decides when to brake, and what evidence would justify pulling the lever.
Why Now
The co-founder did not point to a specific incident that triggered the warning. Instead, they framed it as a broader concern about the current trajectory. With companies racing to release more powerful models, the window for putting safeguards in place is shrinking, they argued.
Past warnings from AI researchers have focused on existential risks — the idea that superintelligent machines could one day destroy humanity. The Anthropic co-founder’s focus on economic instability shifts the conversation to nearer-term dangers. That may resonate more with policymakers who are already debating AI regulation in places like the European Union and the United States.
What Happens Next
No legislative body has yet adopted a brake-pedal model for AI. The debate over how to regulate the industry remains fragmented. The Anthropic co-founder’s comments add a prominent voice to the argument that waiting for a crisis before acting would be a mistake. Whether lawmakers take that advice — and what form their brake pedal might take — is still an open question.
