I’m building a team in DC. Everyone in Washington has a stake in ensuring that reliable info about alignment and RSI reaches the public. We need trustworthy communicators who can explain METR's loss-of-control risk assessments and goals to policymakers.

Some believe that we need work for purpose and dignity, while others see freedom from work as utopic. New paper: What does the *empirical evidence* tell us about work and wellbeing? And what does that imply for AI futures?

No modern conflict has ended by destroying the enemy’s ability to resist—not even World War II. To end a war, both sides must choose to stop fighting, and those choices always require nontrivial bargaining. Misunderstanding the relationship between bargaining and fighting produces systematic failures to integrate negotiating strategy with campaign planning, from Afghanistan to Ukraine to the US war with Iran. New piece in @ForeignAffairs with Stephen Biddle and Yuri Zhukov on the myth of military victory.

There are enough policy opportunities that we need more editors at @IFP! Shape op-eds, reports, and private memos that let us increase the upside (e.g. infrastructure) and lower the downside (e.g. biosecurity) of a dynamist future ($5k referral bonus) ifp.org/opportunity/editor-a…

In August, the Chinese humanoid robot Tiangong Ultra bested Usain Bolt's 100 meter record. These robots still have serious (though sometimes pretty funny) flaws; this one crashed into a wall. But make no mistake. As Pavlo Zvenyhorodskyi and I write in @ForeignAffairs, AI-powered robotics are going to change the physical world. China is investing substantially in these robots while the U.S. lags behind. In the U.S., the top frontier AI companies are betting that they'll be able to automate just about everything humans can do on a computer in a few years. But that leaves a lot: everything that requires real-world interaction. Some people assume that rapid advances in software will easily diffuse into physical spaces. That's a big assumption to make given how important this technology might be. China has invested in AI-powered robotics for several domestic reasons, including creating robotic caretakers. But robotics could just as well become a source of geopolitical advantage in supply chains and in the military. Many robotics companies in China are struggling to make a profit, but there's serious progress in the underlying technology. There significant parallels to China's playbook in EVs and solar. Clearly, the U.S. needs to rapidly build up its robotics industry. But doing so will require a tough balancing act: decoupling is impractical given the deeply integrated nature of the supply chain. One of the key answers will involve working with allies and partners in Europe and Asia who have advantages in areas like industrial robotics as well as novel data sources, such as Ukraine's battlefield data.

There is a lot of debate in the US about whether American foreign policy towards allies and partners can revert to "normal" after Trump. I wrote up a few thoughts on what I'll be watching re: American policy toward India over the next couple of years in the post linked below:

The American AI industry needs a lot more datacenters than it can build at home. If US allies host these datacenters, they might be able to secure access to frontier AI in return. New for @ForeignPolicy with @SamWinterLevy, on getting compute-for-access right:

U.S. allies are in a precarious position, write @SamWinterLevy and Anton Leicht. They are guaranteed neither a share of the economic benefits from developing AI nor reliable access to U.S. AI systems. foreignpolicy.com/2026/10/08…

An interesting proposal for allied scale in AI

Since I started working in AI policy, I've been both surprised and dismayed by how few ties there are between people in the military and people in the AI safety and policy communities. As AI grows more capable, gaps between those closest to frontier AI developments and experts in the domains most affected by rapid AI progress become increasingly consequential. Few of those domains matter more than the military. So I started a Substack today with a post about bridging the divide between these two communities. I think this makes me the first person to ever have a Substack, so I recommend reading and subscribing to it. 1/

“The notion of military victory is a dangerous myth,” write Stephen Biddle, @SamWinterLevy, and Yuri Zhukov. “Terms and bargaining always matter: no war ends unless both sides choose to accept a settlement and stop.” foreignaffairs.com/united-st…

Most countries won't do what it takes to build their own frontier AI. But they don't want to be dependent on foreign AI without a backup. @SamWinterLevy and I argue there's a middle path: AI middle powers could retain 'AI Breakout Capacity'. There's precedent: Japan doesn't have a nuclear weapon, but always retains the capacity to build one on short notice. Middle powers can make strategic choices today to ensure their ability to build frontier systems tomorrow: onshore compute, secure chip purchasing rights, sustain pre-training expertise, and create the right legal basis and private-sector vehicles for a sprint. They'd still have to spend big to build frontier systems later on. But the longer they retain breakout capacity, the longer they can afford to wait: to see if interdependence with America can work, or until their electorate will let them go all-in on an AI moonshot. Until then, the middle path is their wisest choice: hope that U.S. alignment works out, but retain the capacity to deal with being cut off for good. Read more in our latest for @CarnegieEndow's @CEIPTechProgram: carnegieendowment.org/resear…

I'm glad to see this, as I have long thought the scoffing response to sovereign AI was misguided. In the end, your country's sovereignty and territorial integrity is going to depend on having access to the most capable AI systems. Not to have your own route to superintelligence is to decide, now, to be a vassal state in the future.

Most countries can't afford to build frontier AI today but don't want to depend entirely on the US without a fallback. In a new @CarnegieEndow paper, @anton_d_leicht and I argue for a middle path: an AI "breakout posture." Move forward with US interdependence now, but lay the groundwork to build frontier systems later—onshoring compute, retaining pre-training talent, and setting up the legal and corporate vehicles for a sprint if they're ever cut off.

AISI IS HIRING! It was less than a month ago that I became Director. I joined with the belief that AISI is a world-leading organisation, and living proof that government can build things that work. One month in and I'm even more bullish about the vital role @AISecurityInst plays in frontier AI security. I'm astounded daily by the talent, focus and commitment of the brilliant team I get to work with. There has never been a more urgent time to work on frontier AI security - independently, in the public interest. We have a lot of urgent work to do, our team is growing fast, and we need more brilliant people to fill those roles. Our Red Team uses adversarial machine learning techniques to find failures in alignment measures, control monitors, and misuse guardrails. They are massively scaling up all parts of the team: Alignment: job-boards.eu.greenhouse.io/… Control: job-boards.eu.greenhouse.io/… Misuse: job-boards.eu.greenhouse.io/… AI capabilities in cybersecurity and autonomy are advancing faster than ever. Our Cyber & Autonomous Systems team assesses what frontier and open-weight models can really do - across cyber, autonomy and AI R&D. We're hiring a Cyber Security Engineer and Software Engineer to build the evaluations behind that work: CSE: job-boards.eu.greenhouse.io/… SWE: job-boards.eu.greenhouse.io/… Our Human Influence team studies how AI can shift human decisions and behaviour, using everything from RCTs to multi-agent studies. We're hiring an Engineering Lead to help us grow the ambition and pace of that research: job-boards.eu.greenhouse.io/… We're also hiring exceptional Software Engineers at all seniority levels to join our Core Technology team, which works closely with researchers to build performant tools and infrastructure that enable and accelerate AISI's world-leading AI safety research: job-boards.eu.greenhouse.io/…

Very cool job opening at CSET (share widely!): we're hiring a lead for our research team focused on military & intelligence adoption of AI (plus other natsec applications of emerging tech). We're looking for experienced candidates with a strong background in 1+ of military operations, intel analysis, and acquisitions/requirements/budgeting/T&E. You'll be leading a team of fantastic researchers, guiding ongoing work as well as setting new directions for the team. CSET produces independent, rigorous, technically informed research on what the rapidly changing tech landscape means for national security decisionmakers. This is the team that has produced incredible work on Chinese military adoption, US AI decision support, US-China space launch comparisons, and more - I'll link examples in the thread. Apply here by Oct 5 (2 weeks from today): cset.georgetown.edu/job/nati…