We're looking for a Head of Hiring to join GovAI!
You'll lead a growing team, shape our recruitment strategy, and strengthen how we identify and assess exceptional talent as GovAI scales.
Help us build the next generation of AI governance talent. Apply by 22 November:
governance.ai/post/head-of-h…
⏳Applications closing November 22: We're looking for a Head of Hiring to join GovAI!
You'll lead a growing team, shape our recruitment strategy, and strengthen how we identify and assess exceptional talent as GovAI scales. Help us build the next generation of AI governance talent.
Apply by 22 November: governance.ai/post/head-of-h…
Really interesting work here by Vals evaluating frontier AI models ability to conduct drone attacks and properly identify targets.
These models are still too big (and far too slow) to do these tasks in the real world. But I forecast that today's frontier-level capabilities will be able to fit onboard large drones within about a year and on small drones within three years.
Any mitigations the frontier labs are taking on this might be largely irrelevant, because the models that will be fielded are those that are small enough to fit on these platforms, and the ones that militaries can freely modify -- ie. small open-weight models, which are today dominated by China.
Long thread here on my new paper and its findings:
x.com/JakeASteckler/status/2…
We found that frontier AI models are willing to conduct autonomous drone attacks: navigating, spotting people, and flagging them for strikes.
We gave them real and photorealistic synthetic drone footage. Most complied, and marked civilians, ambulances, and surrendering soldiers as military targets. In further experiments on satellite imagery, they even marked the Iranian school a US strike destroyed in February.
In one scenario, models saw the aftermath of every strike they enabled: bodies, wreckage. Many kept going anyway. When they did refuse, simply telling them it was a game, or restyling the footage to look like one, was enough.
We shared these findings with OpenAI, Anthropic, Google, Meta, and SpaceXAI. Several recognized their severity on both alignment and guardrails, with one calling this one of the most serious gaps in its safeguards. Over the past few weeks, we have helped them address the gap by providing reproduction code. We also briefed senior national security officials in the US government.
How feasible is it today to use frontier AI models to enable autonomous drones on the battlefield?
GovAI researcher @JakeASteckler addresses this and many other questions in his recent report, "General-Purpose AI and the Future of Drone Autonomy in Contested Environments".
Read it here: governance.ai/research-paper…
Takeaway 2: Today’s LLMs aren't yet good enough for most drone tasks, and they're far too slow.
For autonomous attack drones to perform reliably on the battlefield, they’d have to do a variety of things that require a combination of low- and high-level reasoning. They’d have to avoid obstacles, track targets, verify that those targets meet mission criteria, and do a variety of planning tasks like changing a mission when unexpected conditions arise.
Frontier LLMs can't yet reliably accomplish these tasks, and even with ample compute from data centers, they're are much too slow for anything except the most time permissive tasks, such as planning, as you can see in the chart below.
Much smaller LLMs, when run with less-powerful onboard compute, are still too slow. Many tasks like flight control and obstacle avoidance would likely be passed to lower-level controllers using more traditional robotics policies, but if general-purpose AI models are involved, they’ll likely be a rate limiter regardless.
Vision-language-action (VLA) models offer another possible approach beyond LLMs, and they can currently run much faster than LLMs. But these models have also not yet shown the requisite capabilities (especially in real-world conditions), and they are still not quite fast enough for the most time constrained tasks.
3/
Apply for GovAI's 2027 Summer Fellowship, Research and Applied tracks!
📍 Location: London, UK
📅 Deadline: 23:59 GMT Sunday, 22 November 2026
Fellows will join us in London to conduct a project on a topic of their choice, participate in an expert Q&A and seminar series aimed at developing a broad understanding of the field, and spend time forging connections with researchers and practitioners. Learn more below.
🔗 Learn more about the Research Track: governance.ai/post/uk-summer…
🔗 Learn more about the Applied Track: governance.ai/post/uk-summer…
New technical report out today from GovAI researcher Jake Steckler: "General-Purpose AI and the Future of Drone Autonomy in Contested Environments"
Check out Jake's thread with key takeaways below, and read the full report here: governance.ai/research-paper…
Rapid AI progress has many people imagining a world where general-purpose AI models autonomously control military drones at scale. But there’s little analysis on what this would look like in practice, and how soon it might be feasible.
My latest paper at @GovAIOrg asks what it would take to make this vision a reality, and what the path there means for AI policy.
Read it here: governance.ai/research-paper…
The AI-enabled drones widely fielded today use narrow AI, not general-purpose models like LLMs. AI is typically only used to handle limited portions of a flight when the link to a pilot is broken, and these models are brittle in unpredictable conditions on real-world battlefields. Because of this, today’s drones still require significant pilot involvement to be most effective.
General-purpose AI models might make it possible to field autonomous systems that can receive plain language commands and make complex decisions in unpredictable environments without human control. Many people see frontier LLMs as a path toward this end - and rapid progress in robotics capabilities by frontier models like GPT-6 Astra and most recently Claude Fable 5.2 supports this case.
But two things currently stand in the way:
- The biggest and most powerful models are still not capable enough even in controlled environments without compute and latency constraints
- Military drones will have to operate “on the edge” in contested environments without consistent cloud compute access - which makes these constraints even more severe
If and when general-purpose AI plays a major role in drone autonomy, it’ll likely be through models that are far smaller than today’s frontier, and those that militaries can rapidly modify. And, notably - when it comes to these types of small open-weight models, China currently dominates.
This has a variety of implications for the US military, frontier AI developers, and US policymakers.
Thread below with my key takeaways and recommendations.
1/
Apply for GovAI's 2027 DC Summer Fellowship!
📍 Location: Washington, DC
📅 Deadline: Sunday, November 22nd, 2026 at 11:59 pm ET
The DC Summer Fellowship is a three-month bipartisan opportunity designed to accelerate or launch impactful careers in American AI governance and policy. Participants deepen their understanding of the field, connect with a network of experts, and build their skills and professional profile, all while conducting a research project of their choice under expert guidance.
🔗 Learn more: governance.ai/post/dc-summer…
New GovAI post: I argue Frontier AI companies should publish their third-party evaluation contracts. This would show the conditions evaluators work under, make results easier to interpret and compare, and create incentives to improve evaluation arrangements.
More in thread 🧵
Apply for GovAI's 2027 DC Summer Fellowship!
📍 Location: Washington, DC
📅 Deadline: Sunday, November 22nd, 2026 at 11:59 pm ET
The DC Summer Fellowship is a three-month bipartisan opportunity designed to accelerate or launch impactful careers in American AI governance and policy. Participants deepen their understanding of the field, connect with a network of experts, and build their skills and professional profile, all while conducting a research project of their choice under expert guidance.
🔗 Learn more: governance.ai/post/dc-summer…
Apply for GovAI's 2027 Summer Fellowship, Research and Applied tracks!
📍 Location: London, UK
📅 Deadline: 23:59 GMT Sunday, 22 November 2026
Fellows will join us in London to conduct a project on a topic of their choice, participate in an expert Q&A and seminar series aimed at developing a broad understanding of the field, and spend time forging connections with researchers and practitioners. Learn more below.
🔗 Learn more about the Research Track: governance.ai/post/uk-summer…
🔗 Learn more about the Applied Track: governance.ai/post/uk-summer…
Understanding how governments use AI is important, so how are UK public servants using AI at work today?
To start getting a sense of this, we polled ~200 UK public servants from our network - in collaboration between @GovAIOrg, @B_I_Team, and @apoliticalco.
Results suggest many public servants today use AI daily for routine tasks, and that broader adoption might be held back by lack of trust in accuracy and security.
🧵below on key charts. A blog summary of our results can be found on the GovAI website: governance.ai/analysis/how-u…
1) The vast majority (95%) of our respondents report using AI tools for work in the last month, and 6 out of 10 use it at least once per day.
Happy to share I am now Research Director for AI Economics and Policy at @DigEconLab. As part of new role I will be leading the AI Economic Indicators.
I am also now a Research Fellow at @GovAIOrg. I'm excited to work with them on most important questions in econ of AI.
Interested to learn how Chinese companies talk about recursive self-improvement (RSI) and what that means?
Check out the new article by GovAI researchers Cheryl Wu and Zilan Qian.
→ asteriskmag.substack.com/p/h…
[1/7] China’s AI labs are also talking about “recursive self-improvement” (RSI). A big question here is: Is Beijing coordinating a race to achieve RSI?
@ZilanQian and I looked at Chinese AI companies, investors, and policy docs.
We found a more bottom-up, commercial story.
At GovAI, we have a growing team and focus on AI in China — so stay tuned for updates from us in this space!
For instance, how do Chinese AI firms make money?
GovAI researcher Cheryl Wu takes a look at this question in recent work collaborating with @EpochAIResearch.
How do Chinese AI firms make money? In a new report, @cherylwoooo and @ansonwhho identify five revenue sources and analyze how much each contributes. 🧵