Anthropic has released an interactive economic scenario explorer that lets anyone model how different levels of AI capability, autonomy, and adoption could reshape the US economy by 2030 — and the range of outcomes its own economists laid out is wide enough to be genuinely unsettling. The tool, built on a technical report titled “Scenarios for our Economic Future,” maps three distinct paths for US GDP, employment, wages, and the split between labor and capital income, depending entirely on how fast AI capability actually translates into real economic adoption over the next four years.
Three Futures, One Model
In the modest scenario, AI’s economic effect looks roughly comparable in scale to the arrival of the internet — US GDP reaches $34.1 trillion by 2030, about 1.6% higher than it would be without AI at all. That’s the scenario where AI remains a genuinely useful productivity tool but doesn’t fundamentally restructure how most work gets done.
The substantial scenario is where things get more disruptive: AI performs roughly half of all knowledge work by 2030, and the economy grows at about twice its normal rate, pushing GDP to $36.3 trillion, 8.3% above baseline. Anthropic’s own survey of nearly 11,000 US adults found that median public expectations cluster closest to this middle scenario — people broadly sense AI is heading somewhere significant, with GDP roughly 10% higher and unemployment climbing to around 5%.
The extreme scenario is the one generating the most attention, and for good reason: annual growth near 15%, GDP reaching $44.4 trillion, alongside historic increases in unemployment and a fall in knowledge-worker pay of more than 10%. Read together with the specific figures buried in Anthropic’s own modeling — unemployment potentially reaching 17.9% and wages for knowledge workers falling as much as 11.5% in that scenario — the tool isn’t just forecasting growth, it’s forecasting who captures it. In every scenario except the mildest, the gains skew heavily toward capital owners rather than the workers whose labor AI is displacing.
Why a Lab Is Publishing Labor Economics, Not Just Model Cards
It’s worth pausing on why an AI company is the one publishing this kind of analysis at all, rather than a government statistics agency or an independent economics institute. Anthropic has increasingly positioned itself as the AI lab most willing to publicly name the disruptive downside of its own product category, a stance that has shaped its public messaging on job displacement for over a year now. Releasing an interactive tool that lets policymakers, journalists, and the public plug in their own assumptions and watch unemployment and wage numbers move is a different kind of communication than a typical corporate research report — it’s closer to an invitation for public debate about outcomes the company itself says it can’t fully control or predict, dressed up as a genuinely useful modeling instrument.
That framing matters because it lands the same week an Anthropic researcher publicly estimated a double-digit probability that advanced AI could cause catastrophic harm to humanity within a decade, following a colleague’s resignation over safety concerns. Taken together, Anthropic’s public output this week reads less like routine product marketing and more like a company trying to get ahead of a narrative it worries is otherwise going to be written entirely by critics, regulators, and displaced workers rather than by the lab itself. The interactive nature of the tool is also a deliberate design choice: rather than publishing one fixed forecast and defending it, Anthropic is inviting economists, journalists, and lawmakers to stress-test its own assumptions in public, which both diffuses some of the criticism any single point estimate would attract and quietly normalizes the idea that a wide range of genuinely disruptive outcomes is on the table.
What This Means for Philippine Founders
This scenario explorer should be read carefully by anyone connected to the Philippine business process outsourcing sector, which employs well over a million Filipinos and remains one of the country’s largest sources of dollar-denominated formal employment. The substantial and extreme scenarios both model roughly half or more of knowledge work being performed by AI by 2030 — and voice-heavy customer support, back-office processing, and entry-level analytical work are exactly the categories most exposed to that shift. Founders building AI-augmentation tools for BPO operators, rather than pure replacement tools, have a real window to position their products as the difference between a Philippine BPO firm managing this transition proactively and one that gets disrupted by a foreign competitor’s AI-native cost structure instead.
The wage-pressure finding is the part worth sitting with longest. If knowledge-worker pay genuinely falls double digits in the more disruptive scenarios, that pressure will be felt first and hardest in markets, like the Philippines, where a large share of knowledge work is delivered as a cost-competitive service to foreign clients rather than owned as proprietary product. Philippine founders and policymakers alike should treat this not as a distant US forecast but as an early warning about the specific wage floor their own labor market competes on — and start building the AI-fluency, higher-value-service positioning, and reskilling pathways now, while there’s still a multi-year runway to adapt before the extreme scenario, if it materializes, arrives.
Share this article