Artificial intelligence could dramatically accelerate US economic growth over the next four years, potentially lifting the size of the economy to around $44 trillion by 2030, according to Anthropic co-founder Jack Clark.
Speaking at the IMF and World Bank annual meetings in Bangkok, Clark said one of Anthropic’s internal economic models shows how rapidly improving AI capabilities could transform productivity across a wide range of industries.
The figure is striking.
US economic output is currently a little above $30 trillion, meaning the scenario would imply an increase of roughly 50% within only a few years.
Clark, however, also warned policymakers that the speed of AI development could create serious risks if international rules and safeguards fail to keep pace.
FCRF Launches CP-FRM to Build India’s Next Generation of Fraud Risk Professionals
Anthropic Sees AI Impact Accelerating Faster Than Expected
Clark said improvements in AI capabilities are happening quickly enough to change the economy faster than many governments and institutions are prepared for.
AI systems are already becoming more capable of writing software, analysing scientific information, handling business tasks and assisting workers across industries.
As those abilities improve, companies could potentially produce more goods and services with the same number of workers.
That is the basic mechanism behind the productivity gains Anthropic expects.
An internal Anthropic economic model cited by Clark suggests that, under an aggressive AI adoption scenario, US output could rise to about $44 trillion by 2030.
The estimate should not be treated as a conventional economic forecast.
It depends heavily on assumptions about how quickly AI improves, how widely companies adopt it and whether the technology produces large productivity gains.
More conservative scenarios could result in much smaller economic effects.
But Clark argued that policymakers should prepare for the possibility that AI develops considerably faster than previous general-purpose technologies.
Productivity Gains Could Spread Across Industries
Most current AI use is concentrated in knowledge work.
Software developers use AI to generate or review code, while businesses use models to analyse documents, write reports and automate customer support.
Clark expects those capabilities to spread further.
AI could accelerate scientific research, drug discovery, engineering and other fields where progress currently depends heavily on skilled human labour.
That could allow businesses to complete work faster and potentially lower the cost of producing new technologies.
The broader economic effect would depend on whether those gains reach industries beyond major technology companies.
The IMF has also identified AI investment as one of the factors supporting global growth this year.
Its latest World Economic Outlook says countries closely connected to the AI supply chain are currently performing better than many economies outside it.
But the Fund has repeatedly warned that the gains could become highly uneven.
Countries lacking computing infrastructure, electricity or skilled workers may benefit much less from the technology.
Clark Warns Faster AI Also Brings Bigger Risks
Clark’s message was not simply that AI will make economies richer.
He also warned that increasingly capable systems could create risks in cybersecurity, biology and labour markets.
Advanced AI can help security researchers find vulnerabilities.
The same capability could also help attackers identify weaknesses and automate cyber operations more quickly.
Biology poses another concern.
Clark said especially strong oversight may be needed where AI systems can assist with biological research because sufficiently capable models could potentially help users design or modify dangerous pathogens.
Anthropic has repeatedly argued that some frontier AI risks cannot be managed through voluntary industry commitments alone.
Clark has called for clear standards governing the most powerful systems and stronger cooperation between governments.
His remarks in Bangkok echoed that position.
AI remains under human control today, he said, but policymakers should not assume that the current balance will remain unchanged indefinitely.
Jobs Remain the Biggest Economic Uncertainty
The effect on employment is one of the hardest parts of the AI growth story to predict.
Higher productivity can create new industries and jobs.
But it can also reduce demand for workers performing tasks that machines can increasingly complete.
AI systems are already capable of handling parts of software development, research, writing, translation and administrative work.
As capabilities improve, more occupations could become exposed.
Clark argued that technological change has historically created new jobs even while eliminating others.
However, the speed of the AI transition could make adjustment more difficult.
Workers may need to retrain faster than during previous technological shifts.
The gains could also become concentrated among companies that own advanced models, computing infrastructure and valuable data.
That is why economists increasingly distinguish between AI increasing total GDP and AI improving living standards broadly across society.
A larger economy does not automatically mean every worker benefits equally.
Policymakers Face an Unusual Timing Problem
The challenge for governments is that economic opportunity and technological risk are developing at the same time.
Regulation that is too restrictive could slow innovation and reduce potential productivity gains.
Rules that arrive too late could leave governments responding only after major problems emerge.
Clark told policymakers that international cooperation will therefore be essential.
AI systems and the companies building them operate across borders, while the consequences of cyberattacks, biological misuse or economic disruption can also spread internationally.
The IMF is increasingly treating AI as both an economic growth driver and a source of financial risk.
Its latest assessments warn that markets are already pricing in large future productivity gains from the technology.
If those gains fail to materialise, expensive AI infrastructure investments and highly valued technology companies could face sharp corrections.
Anthropic’s $44 trillion scenario represents the more optimistic side of that debate.
Whether the US comes anywhere close will depend not only on how powerful AI becomes, but on whether companies, workers and governments can turn that capability into genuine productivity.
What this means for you:
AI could create major economic opportunities, but the $44 trillion figure is a high-growth scenario rather than a guaranteed outcome. Workers and businesses should focus on how the technology changes productivity and specific job tasks, not just headline GDP projections.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
