Anthropic co‑founder Jack Clark warns that artificial intelligence is approaching a stage where it can evolve without human oversight. He urges governments to create “brake” mechanisms, compares AI to the early oil industry, and explains how Anthropic’s self‑coding chatbot Claude could soon be entirely self‑generated. The article also covers the company’s upcoming public listing, its stance on recent U.S. executive orders, and Clark’s advice for a generation facing an AI‑driven economy.
1. A Call for a “Brake Pedal” in AI Development
Jack Clark, co‑founder of the AI research firm Anthropic, told BBC Newsnight that the industry currently operates with a gas pedal but lacks a brake pedal. He explained that the ability to slow the pace of artificial intelligence progress is essential as the technology nears a point where it could develop independently of human direction. Clark emphasized that governments must enact policies that keep control of increasingly powerful AI systems and protect society from unintended consequences.
2. The Need for New Regulatory Frameworks
According to Clark, the world must begin serious contemplation of how to regulate AI. He argued that future rules should give the public confidence that AI systems are safe and reliable. He drew a parallel with the early twentieth‑century oil boom, noting that society responded by establishing sensible policies that allowed the benefits of oil to be realized while limiting the influence of individual industrial leaders. Clark believes a similar approach is required for artificial intelligence.
3. Anthropic’s Self‑Generating Code Milestone
Anthropic’s flagship chatbot, Claude, already runs on code that the system wrote itself for roughly eighty percent of its functions. Clark stated that reaching one hundred percent self‑generated code could be achievable within two years, a development that would have profound implications for the industry. He did not detail how a “brake pedal” could be technically implemented, but he highlighted the urgency of addressing the issue before such capabilities become commonplace.
4. Reaction to the Recent U.S. Executive Order on AI
Anthropic welcomed the executive order on artificial intelligence issued by President Donald Trump. The order is relatively hands‑off, as it does not require AI companies to submit to mandatory safety testing by the government. Testing remains a voluntary activity. Major AI firms, including Anthropic, OpenAI, and Google, have indicated that they will not pause their research programs in response to the order.
5. Preparing for a Public Market Debut
Founded five years ago, Anthropic has experienced rapid growth and is preparing to list on a public stock exchange. The company is expected to become one of the first newly founded AI firms to go public and could achieve one of the most valuable listings in history. Private investors currently value Anthropic at close to one trillion dollars, or seven hundred forty‑five billion pounds.
6. Transparency Over Reputation
Clark explained that Anthropic’s decision to discuss the accelerating capabilities of AI publicly is not intended to enhance its brand with customers. Instead, he wants to inform the world about the internal observations made by companies working with this unprecedented technology. Since its inception by CEO Dario Amodei, Clark, and a small group of executives, Anthropic has positioned itself as an outspoken voice on the potential risks associated with artificial intelligence.
7. Past Disputes and Ongoing Concerns
Anthropic previously entered a public disagreement with the United States Department of Defense over fears that its AI tools could be employed for mass surveillance of American citizens and for autonomous warfare. Clark expressed personal worry for future generations, stating that society must have a serious conversation about the implications of continued AI advancement. He acknowledged both the great benefits and the significant risks that lie ahead.
8. Economic Disruption and the Role of Creativity
Clark warned that AI agents—autonomous bots capable of performing routine tasks—could displace certain jobs. He noted that major technology companies have carried out large‑scale layoffs over the past year, often citing the growing ability of AI tools to replace the work of hundreds or thousands of software engineers. However, Clark believes that individuals who are more creative and generate better ideas may retain an advantage over AI. He observed that Anthropic is now limited more by the ability to generate good ideas than by the engineering capacity to implement them.
9. Advice for Young Professionals
For young people who feel uncertain about an economy increasingly driven by AI, Clark recommends developing a hobby and pursuing a liberal arts education. He emphasized that those who are creative, read widely, and maintain diverse interests will benefit most from the technology. Curiosity, he said, pays dividends in how individuals can harness AI tools for personal and professional growth.
Conclusion
Jack Clark’s remarks underscore a growing consensus that artificial intelligence must be guided by thoughtful regulation and transparent oversight. As Anthropic prepares for a historic public listing and continues to push the boundaries of self‑coding AI, the industry faces a pivotal moment. The call for a “brake pedal” reflects a broader demand for policies that balance innovation with societal safety, ensuring that the benefits of AI are realized without compromising the future of work, privacy, and human creativity.
