On 13 July 2026, more than 200 economists and AI researchers, including 16 Nobel laureates and senior figures from the largest AI labs, published a four-sentence open letter warning that AI job displacement could arrive faster than any prior technological shift. The letter, "We Must Act Now: A Statement on AI's Transformation of the Economy," was released by the Stanford Digital Economy Lab. The story is not that economists are worried; it is that the labs' own leaders are worried, and are saying so on the record.
TL;DR
- 200+ economists and 16 Nobel laureates signed the "We Must Act Now" letter (Stanford Digital Economy Lab, 13 July 2026).
- Industry signatories include OpenAI CFO Sarah Friar, Anthropic co-founder Jack Clark, Google DeepMind chief scientist Jeff Dean, Eric Schmidt, Reid Hoffman, and Yoshua Bengio.
- The letter warns of "large-scale job displacement" from a transformation "larger than the Industrial Revolution" on a shorter timeline.
- Organiser Anton Korinek is on leave at Anthropic; co-signer Ben Bernanke sits on Anthropic's Long-Term Benefit Trust.
- It is a call for policy scaffolding — not a technology critique.
What does the "We Must Act Now" letter actually say?
The letter is only four sentences long. It states that AI could become "radically more powerful over the next 10 years" and could drive "an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame." It flags "major gains in living standards" and "large-scale job displacement" in the same breath, then calls for "incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society."
The brevity is the design. The full text and roster are hosted at wemustactnow.ai, with a companion announcement from the Stanford Digital Economy Lab.
Who signed the letter, and why does the list matter?
The letter was organised by Erik Brynjolfsson (Stanford), Ajay Agrawal (Toronto), Anton Korinek (University of Virginia, on leave at Anthropic), and Tom Cunningham (METR). The signatories fall into three groups that rarely appear on the same document.
The Nobel laureate cohort includes Michael Spence, Daron Acemoglu, Simon Johnson, Joseph Stiglitz, Christopher Pissarides, Paul Milgrom, George Akerlof, Philippe Aghion, Paul Krugman, Ben Bernanke, Roger Myerson, Michael Kremer, and Peter Howitt — a spread across labour economics, growth theory, and monetary policy.
The AI-industry cohort is the unusual part: Sarah Friar (OpenAI CFO), Jack Clark (Anthropic co-founder), Jeff Dean (Google DeepMind chief scientist), Eric Schmidt, Reid Hoffman, Yann LeCun, Yoshua Bengio, and Dan Hendrycks. These are the people shipping the models, agreeing on the record that their technology could cause large-scale job displacement.
Two adjacencies sharpen the point: Anton Korinek, one of the organisers, is on leave at Anthropic. Ben Bernanke, a Nobel signatory, was recently appointed to Anthropic's Long-Term Benefit Trust. An AI lab's own economist and one of its trust members are among the loudest voices asking policymakers to plan for job losses.
Why is "AI job displacement" different from past automation waves?
Every organiser makes the same core argument: prior general-purpose technologies gave societies decades to adapt; this one may not. Anton Korinek: steam, electricity, and computers each afforded decades while AI "may give us only a few years." Cunningham described the visibility as "driving in the fog." Brynjolfsson framed the gap between capability growth and economic understanding as "the greatest opportunities of our era."
The mechanism is speed of substitution across cognitive tasks. Earlier automation waves displaced occupations slowly enough for labour markets to reshuffle. If frontier models compress that window, standard retraining timelines stop working. Acemoglu frames the response as redirecting AI "so that its risks are minimized and it can work for the benefit of workers and society."
The letter does not commit to a specific probability or timeline. That reticence broadens the tent: signatories disagree on how fast capabilities will scale but agree institutions are not ready.
What is the letter calling for in practice?
The text stops short of prescribing policies. It asks for "incentives, guardrails, and institutions" that steer AI toward complementing human work. Four workstreams are implied:
- Labour-market instrumentation. Higher-frequency data on task-level automation and displacement.
- Complement-oriented incentives. Tax or procurement rules that reward AI uses which augment workers over replacement.
- Transition infrastructure. Portable benefits and faster reskilling for a shorter adjustment window.
- Governance capacity. Standing bodies with technical fluency to monitor deployment and update rules.
Yoshua Bengio captures the political framing: "We must be intentional and make collective, democratic choices, rather than letting market forces play out and risking leaving most citizens behind." Spence's phrase — "all hands on deck" — is a request for coordination, not a moratorium.
How should operators and executives read this signal?
Treat the letter as a leading indicator for regulation and procurement rules. When the CFO of the largest closed-model lab and the chief scientist of the second-largest research group sign the same document as Nobel laureates in labour economics, the space of politically acceptable AI policy has already shifted.
For workforce planning, separate two questions: where does AI reduce headcount for the same output, and where does AI change what a role produces? The second is more durable. Our Bending Spoons SEC filing review shows how quickly output-per-employee can move when AI is a first-class input.
For cost planning, lab economics and labour economics are converging. Our token efficiency playbook and Reverse Information Paradox analysis argue that firms extracting durable value from AI price the technology honestly and protect proprietary context. The same applies to workforce forecasts: assume neither zero displacement nor total automation.
What are the limitations of the letter?
Three caveats. First, four sentences cannot resolve the empirical disagreements between signatories; the letter is a coordination device, not a research finding. Second, senior lab affiliations do not equal firm-wide positions — Friar, Clark, and Dean sign as individuals. Third, the letter is silent on who bears the transition cost.
FAQ
Q: When was the "We Must Act Now" letter published, and by whom? A: Released 13 July 2026 by the Stanford Digital Economy Lab, organised by Erik Brynjolfsson, Ajay Agrawal, Anton Korinek, and Tom Cunningham.
Q: How many economists and Nobel laureates signed? A: 200+ economists and AI researchers, including 16 Nobel Prize laureates in economics.
Q: Which AI-lab leaders signed? A: OpenAI CFO Sarah Friar, Anthropic co-founder Jack Clark, Google DeepMind chief scientist Jeff Dean, Eric Schmidt, Reid Hoffman, Yann LeCun, Yoshua Bengio, and Dan Hendrycks.
Q: Does the letter predict how many jobs will be lost? A: No. It warns of "large-scale job displacement" alongside "major gains in living standards," without a specific figure or timeline.
Q: What does the letter want policymakers to do? A: Build "incentives, guardrails, and institutions" that steer AI to complement human work rather than replace it — before the transformation is fully under way.
Q: Why is Anton Korinek's role significant? A: Korinek co-organised the letter while on leave at Anthropic — an AI-lab economist publicly warning about AI-driven job losses.
Discussion
0 comments