Could artificial intelligence (AI) make the US economy much bigger in just a few years? Anthropic co-founder Jack Clark thinks it is possible. Speaking in Bangkok on Saturday, October 17, 2026, he told finance leaders that one of the company’s internal models shows AI could boost US GDP to about $44 trillion by 2030 in an extreme case. He also asked governments to act fast on rules. Here is what he said, in simple words.
Where and Why Clark Spoke
Clark spoke at the annual meetings of the International Monetary Fund (IMF) and the World Bank in Bangkok. The audience included finance ministers, central bankers, and other policymakers. He is Anthropic’s head of public benefit and a former journalist. His main message was that AI is improving quickly, change will come faster across every field, and the time to prepare is getting shorter.
The Big Number: $44 Trillion
Clark said an internal Anthropic model projects that US output could reach about $44 trillion by 2030 in its highest scenario. The US economy today is a little over $30 trillion. Clark admitted that the number “sounds very outlandish,” but said it follows if the technology keeps getting better. He added that if GDP grew this much, new industries and inventions would have to appear to power such growth.
Three Possible Futures
Reports say Clark showed three scenarios for US GDP in 2030, measured in constant 2025 dollars:
| Scenario | US GDP in 2030 | Growth Shown |
|---|---|---|
| Modest | $34.1 trillion | 1.6% |
| Substantial | $36.3 trillion | 8.3% |
| Extreme | $44.4 trillion | 32.4% |
Reports describe the size of the jump in different ways. Reuters’s first version said “almost 50%” by 2030 compared with today, while another outlet lists the extreme case as a 32.4% boost. The two numbers likely use different starting points, so check Anthropic’s own materials before you quote a percentage. One report also says the extreme case needs AI systems that can improve themselves over and over.
Why Clark Thinks AI Could Speed Up Science
Clark said AI is vastly increasing scientific discovery. He described a future where human researchers work with hundreds or thousands of AI assistants. These helpers could solve problems in mathematics, engineering, and biology, including finding new medicines.
The Risks He Pointed Out
- Misuse of science: Clark said using AI for science also opens the door to misuse, such as biological and cyber weapons, and to things we cannot yet imagine.
- Biology needs the strictest care: He said biology is an area where the strictest human oversight and openness are needed to keep people safe. He added that AI systems have already created new viruses.
- Losing control: He said there is growing evidence of a possible “singularity,” a point where AI development moves faster than humans can control.
- Jobs: He said some jobs with “high AI exposure,” such as call centre and customer service work, are already being lost.
Officials at the meeting also worried that AI could destroy certain jobs and leave some countries behind.
What He Asked Governments to Do
Clark said countries need to work together to create “unusual and serious rules” for this powerful technology. He warned that the window for policymakers to act is shrinking, and that leaders must understand how strong AI could become before making choices about rules.
The Mood in the Markets
His talk came during a rough stretch for AI-linked investments. One report says US chip companies lost about $1.64 trillion in value across six trading sessions, as investors worried about high prices, borrowing costs, and whether AI spending will pay off.
Why This Matters
These numbers are a projection from one company’s model, not a prediction that is certain to happen. Even Clark called the top number outlandish. Still, the speech shows how seriously AI companies and economists now take the question of fast AI progress. For governments, the message is to plan for several outcomes, from slow growth to very fast change, and not just one.
Frequently Asked Questions
What did Jack Clark say about US GDP?
He said an Anthropic internal model shows US GDP could reach about $44 trillion by 2030 in an extreme case, compared with just over $30 trillion today.
Is $44 trillion Anthropic’s main forecast?
No. It is the highest of three scenarios. The other two are about $34.1 trillion and $36.3 trillion.
Where did he speak?
At the IMF and World Bank annual meetings in Bangkok on October 17, 2026.
What risks did he mention?
Misuse of AI for biological and cyber weapons, loss of human control, and job losses in areas such as call centres and customer service.
What did he ask policymakers to do?
Work together on serious international rules and act soon, because the window to act is getting smaller.
Clark’s message to policymakers was simple: AI is moving fast, the possible gains are huge, and so are the risks. Whether the extreme $44 trillion case ever happens, governments are being asked to prepare now.
