1/ The Pacing the Frontier letter calls on the US government to support an international effort to build the technical and governance tools needed to protect our option to pace AI development. bsky.app/profile/yosh...

Yoshua Bengio (@yoshuabengio.bsky.social)

1000+ scientists at frontier AI companies are speaking out to warn that the current commercial race leads to unacceptable security risks. I agree with their call for an international effort to…

bsky.app

1/ Air Canada had to honor a discount its chatbot invented. The liability caused by AI agents is landing on policies written by an insurance industry that never planned for them.

1/ AI agents that can sign contracts on your behalf, hire employees, set prices, and move your money around are being heavily invested in by AI companies. But what I want to call attention to, what happens if an agent sells you faulty goods or runs off with your deposit?

I joined over 200 economists and AI researchers in signing "We Must Act Now" a statement on AI's transformation of the economy. AI could reshape the economy at unprecedented speed. The opportunities are enormous and so are the challenges. We need to start preparing our institutions.

We Must Act Now: A Statement on AI’s Transformation of the Economy

wemustactnow.ai

1/ Illinois just became the first state to require frontier AI developers to undergo annual third-party audits. Gov. Pritzker signed the AI Safety Measures Act (SB 315) this week, going beyond California and New York, which only require published frameworks and incident reports.

1/ Hundreds of contractors posed as teenagers to test how rival chatbots like ChatGPT and Gemini handle prompts about suicide and self harm. The project was run for Meta. buff.ly/Auslrgh

Meta Contractors Posed as Teens to Prompt Rival Chatbots About Suicide, Sex, and Drugs

Hundreds of contractors working on a project for Meta pretended to be kids in order to see how other chatbots like Gemini and ChatGPT would respond to high-risk subjects, WIRED found.

wired.com

So thrilled for Karolina Stańczak, presenting our paper at FAccT today. You cannot write a complete contract for an AI, and we never wrote one for ourselves either. Norms, markets, and law fill that gap, and the paper asks what alignment could learn from them. Read here: arxiv.org/abs/2503.00069

Societal Alignment Frameworks Can Improve LLM Alignment

Recent progress in large language models (LLMs) has focused on producing responses that meet human expectations and align with shared values - a process coined alignment. However, aligning LLMs…

arxiv.org

Most AI safety work tests one model at a time. Put many agents together and you get cooperation but also risks like collusion and cascades. A new Schmidt Sciences call, with @aria-research.bsky.social, the @coop-ai.bsky.social (whose board I chair), and others. Open globally, due Aug 9.

A few details for anyone on the fence. We're aiming for a cohort of 20 to 25 fellows and we're reviewing applications as they come in. Earlier is better, but get yours in by July 15 to be assured of full consideration. bsky.app/profile/ghad...

Gillian Hadfield (@ghadfield.bsky.social)

Applications are open for a new AGI Governance Fellowship at Johns Hopkins, led by @sethlazar.org, Nicholas Caputo, and me. Three weeks in DC this September, in person, for early-career people…

bsky.app

Applications are open for a new AGI Governance Fellowship at Johns Hopkins, led by @sethlazar.org, Nicholas Caputo, and me. Three weeks in DC this September, in person, for early-career people already in AI governance. A capstone for the next generation, not an entry point.

My new op-ed is out in the Washington Examiner. The Great American AI Act would have the most powerful AI models tested for catastrophic risk by independent, government-licensed verifiers, with the government, not the companies, setting the standard. buff.ly/G45GK0Q

The uncomfortable truth behind AI debate — and a bipartisan solution

The Great American AI Act from Obernolte and Trahan builds on a model for regulating complex, fast-moving technologies that my colleagues and I developed.

washingtonexaminer.com

My colleagues at Fathom have a sharp new piece: independent verification has gone from a proposal to a live national debate. I've been working on this since 2017. This shift is real progress, but only the first step, and the harder question is what it should look like.

Independent Verification Goes Mainstream

A bipartisan federal discussion draft, Anthropic, OpenAI, and Google DeepMind are all converging around independent verification as a key element of AI governance.

fathomai.substack.com

1/ Grateful for Dean Ball's sharp and balanced take on politics and AI governance, and the crazy situation we've landed in: increasingly urgent to get a smart regulatory framework for powerful AI, and yet basic rule of law and common sense are getting kicked to the curb. buff.ly/JfV0BiN

Leviathan Waking

On Anthropic/USG, and a new era in AI governance

hyperdimensional.co

In a new essay, Anthropic CEO Dario Amodei calls for mandatory third-party testing of frontier AI models, with the government empowered to block unsafe deployments. One way to do it, in his words: a regulatory markets approach. His co-founder Jack Clark and I proposed that in 2019. buff.ly/rmhlHZn

Dario Amodei — Policy on the AI Exponential

In one of the side plots to The Lord of the Rings, two of the Hobbits attempt to rouse Treebeard—a wise but ponderous sentient tree—to defend his forest from an army that is cutting it down. The…

darioamodei.com

New interview with Sinead Bovell on I’ve Got Questions. My version of existential risk for AI isn’t rogue superintelligence. It’s releasing billions of AI agents into the economy before we have legal infrastructure for identity, registration, or liability.

Could AI Agents Crash the Economy? | Gillian Hadfield (AI Economist)

Are we looking at the beginning of the end of the internet as we know it? In this episode of I’ve Got Questions, I sit down with Professor Gillian Hadfield, a leading scholar in AI alignment,…

youtube.com

1/ The companies building AI are making consequential value choices every day. Right now, they are making those choices for the rest of us. New piece with @AndrewFATHOM on @Fathom_org Substack.

Thanks to Anton Korinek and UVA’s EconTAI Initiative for hosting me today to present work with Rakshit Trivedi and Dylan Hadfield-Menell on building AI agents that can participate in democratic life without destabilizing it. Paper at the Knight First Amendment Institute:

Building AI for the Democratic Matrix: A Technical Research Agenda for Normative Competence and Normative Institutions

To maintain democratic resilience, it is essential to build AI agents capable of choosing behaviors that mirror those of the human agents that constitute human democracies.

knightcolumbia.org

@deanwb.bsky.social in The Economist makes the public case for an architecture Dean and I have both been advancing: a network of independent organizations that audit AI safety claims. Worth a read. buff.ly/gZZCh8D

No to laissez-faire on AI, yes to a light touch

The private sector can do much of the heavy lifting in verifying safety claims, writes Dean Ball

economist.com

The second TAIGR workshop at ICML is open for submissions through April 24. Full papers and mini papers both welcome.If you're working on technical AI governance, send it in. If you know someone who is, pass this along. taigr-workshop.com

TAIGR @ ICML 2026 — Workshop on Technical AI Governance Research

Second Workshop on Technical AI Governance Research at ICML 2026. Bridging ML researchers and policymakers in Seoul, South Korea.

taigr-workshop.com

1/ Who decides how AI systems behave? With Tianmin Shu, @sethlazar.org, and @dhadfieldmenell.bsky.social, we were named runner-up in the LaudeInstitute's inaugural Moonshot program and awarded a seed grant to find out. www.laude.org/moonshots

Laude | Moonshots

We asked the most consequential AI researchers in the world how they would use AI to solve humanity's hardest problems. 125 proposals and 600 researchers later, meet the Moonshots // ONE awardees.

laude.org

1/ Something I’ve been working toward for a long time. Virginia just signed the first state legislation directing a formal study of Independent Verification Organizations for AI. Bipartisan votes of 84-14 in the House, unanimous 40-0 in the Senate.

1/ Social media makes it look like the public is deeply divided. But what if most of the public just isn’t speaking? In a new paper with Atrisha Sarkar in JAIR, we show why. We call it rational silence.

Democracy isn't a rulebook. It runs on daily interactions where people comply with norms and hold each other accountable. AI agents are about to join that system. We need to build them to read it. New paper with Rakshit Trivedi and Dylan Hadfield-Menell.

Building AI for the Democratic Matrix: A Technical Research Agenda for Normative Competence and Normative Institutions

To maintain democratic resilience, it is essential to build AI agents capable of choosing behaviors that mirror those of the human agents that constitute human democracies.

knightcolumbia.org

At the @FAR.AI London Alignment Workshop I made the case for Independent Verification Organizations: licensed, competing private entities that can grow our regulatory capacity to match the pace of AI capabilities. Full talk: bsky.app/profile/far....

FAR.AI: Frontier Alignment Research

FAR.AI is an AI safety research non-profit facilitating technical breakthroughs and fostering global collaboration.

far.ai

AI systems are quickly becoming embedded throughout the economy. But we have almost none of the regulatory tools, regulatory markets among them, to manage them. Here's what I think we should do about it: www.americanbar.org/groups/scien...

Regulatory Markets: The Future of AI Governance

Regulatory markets can bridge technical and democratic gaps in AI governance by pairing public oversight with private, licensed regulatory innovation.

americanbar.org

“The most practical governance framework currently in circulation.” That’s Forbes on the Independent Verification Organization model Fathom and I have been developing. Legislation takes years; IVOs move at the pace of innovation.

"Why not work on what kind of new governance is needed to ensure secure, reliable, predictable use of all frontier models, from all companies?"

In London today and tomorrow for the Alignment Workshop organized by FAR.AI. Keynoting alongside Rohin Shah and Allan Dafoe. I look forward to seeing everyone in attendance! www.far.ai/events/event...

FAR.AI: Frontier Alignment Research

FAR.AI is an AI safety research non-profit facilitating technical breakthroughs and fostering global collaboration.

far.ai

The 2026 AI Safety Report's biggest finding isn't the risks it catalogs. It's the evidence gap. We're trying to build AI governance with almost no science underneath. Massive investment in the research regulatory systems depend on is overdue. internationalaisafetyreport.org

International AI Safety Report

The International AI Safety Report is the world's first comprehensive review of the latest science on the capabilities and risks of general-purpose AI systems. The work was overseen by an…

internationalaisafetyreport.org

Why not work on new governance...