Should governments act on AI now?
Who says what
Researchers’ documented positions on “Should governments act on AI now?”
Documented positions only, coded with cited evidence by an AI agent: 26 of 400 mapped researchers so far. Everyone else is unknown: no reviewed evidence, which is neither agreement nor neutrality.
Agree (23)
Anton Korinek agrees
Argues society should actively steer technological progress rather than accept its default direction.
Evidence
- Steering Technological Progress
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Daniel Trefler agrees
Has published an economic case for regulating artificial intelligence and discusses its policy implications, alongside research on how cross-border data regulation shapes AI-driven digital trade.
Daron Acemoglu agrees
Argues technology's direction is a policy choice and calls for actively steering AI toward pro-worker uses now.
Evidence
- Building Pro-Worker Artificial Intelligence
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Gillian Hadfield agrees
Legal scholar whose research program designs AI governance — regulatory markets, compute governance, and international institutions — arguing current legal infrastructure is not ready.
Jakub Growiec agrees
Argues the existential risk from transformative AI calls for more public funding of AI safety research and stronger AI regulation, ideally coordinated globally.
Joseph Stiglitz agrees
Argues AI's direction cannot be left to the market and that government regulation may be needed, and his recent work proposes deliberately steering innovation toward labor-complementing technologies.
Evidence
- “One can’t just leave it to the market.” — source
- “We may have to use government regulation because of the weakness of the bargaining power of workers” — source
- Steering Technological Progress
Yoshua Bengio agrees
Leading advocate for AI governance: chairs the International AI Safety Report and argues for compute governance and international institutions for advanced AI.
Abhishek Nagaraj leans agree
His co-authored work on generative AI competition argues that tight control of complementary assets could concentrate the market and suggests policy interventions forcing shared access to those assets may help preserve competition; it does not address AI regulation broadly.
Christian Catalini leans agree
Calls on policymakers to treat verification infrastructure as a public good and to implement liability regimes, incident reporting and auditable execution traces.
Evidence
- “Policymakers must treat verification infrastructure and ground truth as public goods.” — source
Christopher A. Pissarides leans agree
The Pissarides Review final report recommends building policy and regulatory architectures to monitor automation and steer it toward better job quality; he has not called for specific AI legislation in sources found.
Evidence
- “we suggest building practical, policy and regulatory architectures for ongoing monitoring” — source
Daniel Susskind leans agree
Writes that new evidence on declining labor demand is significant for AI governance and names the growing political power of large technology companies as a problem to address.
Evidence
- Technological Unemployment
- “the growing political power of the large technology companies” — source
David Autor leans agree
Argues for actively building pro-worker AI rather than leaving the direction of the technology to default incentives.
Evidence
- Building Pro-Worker Artificial Intelligence
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Diane Coyle leans agree
Co-authored policy work arguing for active government strategy on AI, including compute-based governance, an integrated UK approach to AI sovereignty, and public-value-oriented AI development.
Emilio Calvano leans agree
Co-authored work finds AI pricing algorithms can learn to collude tacitly and calls for fresh policy approaches to AI.
Erik Brynjolfsson leans agree
Calls for urgent research and institutional preparation for transformative AI's economic disruption.
Evidence
- A Research Agenda for the Economics of Transformative AI
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Georgios Petropoulos leans agree
He calls for market regulation, competition policy and labour-market policy for AI-related markets and served on the expert panel advising on the EU Digital Markets Act; in 2017 he cautioned against rules adopted without understanding the technology.
Evidence
- “Addressing the market-power failure in AI-related markets would require a combination of market regulation, competition policies and labour-market policies” — source
- The EU Digital Markets Act: A report from a Panel of Economic Experts
Giacomo Calzolari leans agree
Co-authored work finds AI pricing algorithms can learn to collude tacitly and calls for fresh policy approaches to AI.
Klaus Prettner leans agree
Studies optimal taxation of AI and policy challenges 'in the race against the machine' — policy action is his research premise.
Martin Beraja leans agree
Shows market-driven automation can be socially inefficient and studies when governments should intervene in AI development.
Michiel Bakker leans agree
Argues Europe needs an AGI preparedness plan — institutional action ahead of capability arrival.
Pascual Restrepo leans agree
Argues the direction of AI research is a social choice that policy should influence.
Evidence
- The wrong kind of AI? Artificial intelligence and the future of labour demand
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Philippe Aghion leans agree
Argues competition and labor-market policy shape whether AI's growth benefits materialize.
Evidence
- Artificial Intelligence and Economic Growth
- “Signed the wemustactnow.ai open letter urging policymakers to prepare for AI's economic impact” — source
Robb Willer leans agree
His experiments find that LLM-generated messages can shift people's policy attitudes, and he has said this raises high-stakes questions for lawmakers that should be addressed early; his stated position concerns AI in political persuasion rather than AI regulation generally.
Evidence
- LLM-generated messages can persuade humans on policy issues
- “AI has the potential to influence political discourse, and we should get out in front of these issues from the start.” — source
It's complicated (3)
Ajay Agrawal it's complicated
His policy work weighs supporting AI diffusion against premature restriction — policy matters, but heavy-handed regulation has costs.
Arvind Narayanan it's complicated
Supports evidence-based governance but pushes back on restrictive rules driven by speculative risks, e.g. defending open foundation models.
Joshua Gans it's complicated
Argues AI policy should balance enabling diffusion against specific harms, warning against reflexive restriction.