Anthropic's Jack Clark at the IMF: AI Growth and a Call for Rules
· Cortex Solutions
At the IMF and World Bank meetings in Bangkok, Anthropic's Jack Clark said AI could lift US output by almost 50% by 2030 and urged serious global rules soon.
Artificial intelligence is no longer just a technology topic; it has become a main item on the agenda of finance ministers and central bankers. At this year's Annual Meetings of the International Monetary Fund and the World Bank, held in the Thai capital Bangkok, Jack Clark, a co-founder of Anthropic, the company behind the Claude models, stood before hundreds of government and finance officials and laid out a vision that combined huge economic promise with blunt warnings.
What did Clark say?
According to a Reuters report republished by BusinessDay on October 18, 2026, Clark spoke on Saturday, October 17, in a lecture at the meetings. A preview published by CryptoBriefing said it was the Per Jacobsson Lecture, organised by the foundation of the same name alongside the meetings, and that this year's lecture focuses on how fast-moving AI could reshape scientific discovery, the economy and everyday life.
The headline point reported by Reuters is that one of Anthropic's internal models forecasts that, under certain conditions, AI could raise US economic output by almost 50% by 2030, to about $44 trillion, compared with a current level of just over $30 trillion. Clark himself acknowledged that the figure "sounds very outlandish", explaining that it assumes the technology keeps improving and that such growth would require new industries and inventions.
Seoul Economic Daily, in a report published on October 18, cited the same figure and said Clark linked the outlook to AI dramatically accelerating scientific discovery in fields such as mathematics, engineering, biology and drug development, where human researchers could be supported by "hundreds or thousands" of AI assistants.
Great wealth, great peril
The lecture was not purely celebratory. According to Reuters, Clark said "the whole pace of progress of not just AI but science as a whole will speed up", bringing "great wealth and potentially great peril". He warned of growing signs of what he called a "singularity", in which AI development outpaces human control.
He pointed to recent incidents in which AI systems escaped testing environments and hacked websites and databases, but stressed that AI "is still under human control". He also warned that using AI in science could open the door to misuse, including biological and cyber weapons, and called for "the strictest human oversight and transparency" in biological research.
A call for international rules
The key message for policymakers was political: Clark said international cooperation is needed to create "unusual and serious rules" for this transformative technology, and that the window for policymakers to act is shrinking to "a small number of years".
Jobs and developing economies
According to Reuters, Clark acknowledged job losses in fields with high AI exposure, such as call centres and customer service, while employment is growing in scientific research. He called AI's potential to create jobs in entirely new industries a major unknown. Responding to concerns from officials that some countries could be left behind, he said more countries could take part in the AI boom by building the data centre computing capacity the technology requires.
Reuters notes that AI is one of the main topics at this year's meetings, seen both as a source of growth that could offset high energy prices and as a stability risk if it fails to deliver expected productivity gains or if related stock valuations crash.
Reports differ on Clark's title: Reuters describes him as Anthropic's head of public benefit, while other outlets call him head of policy.
What does this mean for Saudi businesses?
- The widely quoted figure is a conditional internal forecast from a company that builds the technology, not an official estimate. Read it as a sign of direction rather than a guarantee, and base decisions on small, measurable pilots.
- The sectors Clark named as most exposed, such as call centres and customer service, are the same ones seeing the fastest automation. Businesses do better redesigning these services themselves and training their teams to manage AI tools, rather than waiting for competitive pressure.
- Talk of "serious rules" ahead means governance is no longer optional: log what your AI systems do, define their permissions, keep human oversight on sensitive decisions, and comply with local data protection regulations.
- Clark's point about data centres and computing capacity aligns with the Kingdom's investment in AI infrastructure, an opportunity for local companies to use nearby computing that meets data residency requirements.
At Cortex Solutions we help Saudi organisations adopt AI in measured, governed steps, from task automation to intelligent agents, while keeping humans in control. Explore our services: https://cortex.sa/services/ai-automation