Anthropic Co-Founder Says AI Industry Lacks a Brake Pedal
BBC Business reported Thursday that Anthropic co-founder Jack Clark is urging governments to develop tools to slow artificial intelligence development before the technology advances beyond meaningful human control.
Speaking to BBC Newsnight, Clark said the AI industry currently operates with only an accelerator and no mechanism to decelerate. He argued that policy frameworks, not corporate goodwill, must ultimately keep increasingly powerful AI systems in check.
A Warning From Inside the Industry
Clark’s concern centres on the pace of autonomous AI development. He noted that Anthropic’s flagship chatbot Claude already generates roughly 80% of its own code. Clark told the BBC that reaching full self-authored code is plausible within two years and would carry profound consequences for how the technology is governed and deployed.
He stopped short of prescribing a specific regulatory structure. Instead, he drew a historical comparison to the early oil industry, arguing that society eventually built frameworks that let people benefit from oil without depending on the character of those running oil companies. Clark suggested AI governance needs to follow a similar path.
Also Read: Trump Signs Executive Order on AI Development
Background: Anthropic’s Safety-First Positioning
Anthropic was founded in 2021 by Clark alongside chief executive Dario Amodei and five other former OpenAI researchers. The company has consistently presented itself as more safety-conscious than its rivals. That stance was tested recently when Anthropic publicly challenged the US Department of Defense over concerns that its tools could enable mass surveillance or autonomous weapons systems.
The company’s warnings also arrive in a commercially consequential moment. Anthropic is preparing a public stock listing that private investors value at close to $1 trillion, which would rank it among the most significant market debuts in history. Clark insisted the public commentary on AI risk is not a marketing exercise.
Also Read: Anthropic Plans US Share Sale as Valuation Nears $1 Trillion
Jobs, Creativity, and What Comes Next
Clark acknowledged widespread anxiety about AI-driven job displacement. He said AI agents, autonomous bots handling routine tasks, could displace workers across many sectors. Major technology companies have already cited AI capability as a rationale for large-scale redundancies over the past year.
Yet Clark offered a measured counterpoint on human creativity. He argued there is no solid evidence that AI systems can genuinely innovate, and said Anthropic is now constrained more by the supply of good ideas than by engineering capacity. His advice for younger workers entering an AI-shaped economy was direct: pursue broad intellectual interests, read widely, and lean into curiosity.
The Trump administration’s recent executive order on AI notably left safety testing voluntary, a position Anthropic endorsed despite Clark’s call for stronger regulatory guardrails.
Read Next: What Is Claude and Why Is Anthropic’s Chatbot Attracting Attention?
