How can AI support public defense? We have worked on this question for the last two years, and have (1) built an AI retrieval tool for the NJ Office of the Public Defender and (2) conducted qualitative interviews with defenders. Our main insights 👇(1/12)

This is wild, but really great investigative journalism. I guess all the jailbreaking papers trying to prevent "How do I build a bomb?" queries have been vindicated... So much for it being an unrealistic scenario... www.nytimes.com/2026/07/10/u... casp.ac/reports/ai-e...

You can also check out our website and explore the work in an interactive way! Pick which question you think is better from justices or models, see the coverage of different legal issues from various models, & more! princeton-polaris-lab.github.io/oral-args-we...

"Altman told staff that the government would be 'approving access customer by customer during this preview period'" Opaque de-facto licensing regimes are a recipe for corruption. Not sure this is the best place to have landed on AI governance... www.theinformation.com/articles/tru...

The future of oral argument or trial prep will obviously involve AI. But how can we measure quality of simulations? We build out an eval suite and run some studies! Sycophancy & issue coverage are clear areas of improvement, but models are surprisingly useful! 👇👇

A surprisingly detailed accounting of how Grok is used in DoW, presumably along with other models in Maven. Maven workflows "enabled" US forces to hit 2000 targets in Iran over 96 hours. storage.courtlistener.com/recap/gov.us...

The Fable export control is just one potential method that the government can use to pull models offline. The gov could also eventually use the Atomic Energy Act, saying that a model is "born secret" if it can reconstruct classified nuclear secrets. Check out blogpost on it!👇

The rule against viewpoint discrimination is one of the most imptl in First A law. But as I show in a new paper, forthcoming in the U Penn Law Review, the test of viewpoint discrimination has changed a LOT in the past few decades, in good ways and bad. 🧵 papers.ssrn.com/sol3/papers....

Forcing people to leave the country to transfer to a green card, will hurt families, communities, and United States innovation. If you aren't aware, adjustment of status can take many months—if not years. I hope this policy is reconsidered.

Current discussions on AI's labor impacts seem to glaze over: How long will the transition take, in our lifetime? Will wages be the same? Will power be more/less concentrated? Will peoples' new jobs be as satisfying as the career they had built? Will economic mobility be the same?

Check out new semantic search on CourtListener which @dominsta.bsky.social and our lab contributed to!

if you try to get Claude to speak Armenian it just outputs "delays"! Seems like glitch tokens are still unresolved. Interesting (kind of sad?) to see Opus thrown into a loop.

From oral arguments yesterday!

Btw, did a bit of a rebranding of the substack. Will endeavor to post more there. h/t @dbateyko.bsky.social on the Trials & Errors name. Super fitting name for a group whose focus is both in reinforcement learning and in law/governance research. www.trialserrors.ai

This is a challenging legal problem for NeurIPS (and other conference participants)! You might be wondering how this is possible given the First Amendment? I wrote a quick explainer on the current status quo of relevant First Amendment cases & law to get you up to speed. 🔗👇

I feel this urgency too. But this is all so utterly avoidable with good policymaking. No one should be left behind because they didn't accumulate capital in 2026. There are so many people who aren't plugged into these conversations or are simply not in a position to do anything about it.

I’m really excited about our new paper! I think we will ultimately need to draw on expertise from both law and AI to get alignment right, and this paper lays out that vision in more detail. arxiv.org/abs/2601.04175

The current direction of AI labs is “we’re building something that’s going to replace you and we have no plan to make sure you’re going to land in a better place, but we’ll make billions.” The logical reaction is, “shut it down.” Labs need to get serious on addressing labor impacts.

Many legal scholars talk about lock-in effects for LLMs from the conversation history/memories (akin to social media). But if an LLM can access the info & is capable, you can just ask the LLM to give it to you, making it far easier to switch providers than social media. Good example of that here.

Only a couple of days after my last post, vibe hacking in full force.

Only a couple of days after my last post, vibe hacking in full force. www.bloomberg.com/news/article...

Following by a panel on GenAI, Agentic AI, Law, and CS (1:15-2:00pm ET) with @peterhenderson.bsky.social (Princeton) and Georgios Piliouras (Google DeepMind) Spotlight Talks (2:30pm-4:00pm) by @aloni-bologna.bsky.social (UChicago), Rebecca Wexler (Columbia), and @jubaz.bsky.social (Georgia Tech)

Missing from the headline: "using Claude Code." Vibe hacking is already a thing. I've been saying this for a while, but no model-level safeguards will prevent it entirely. What they can do is slow it down enough for us to put societal-level safeguards in place. www.popsci.com/technology/r...

That was fast.

New copyright law "hypothetical" just dropped.

Warner Music and Udio settle their copyright case, agree to collaborate on "new song creation service that will allow users to remix tunes by established artists." Expect more such settlements as copyright holders look to leverage AI to boost revenue!

We’ve been pushing hard on AI for public good. One example: partnering with Courtlistener to launch accessible legal semantic search! Many more cool AI projects coming soon from my group aimed at improving access to justice, often spearheaded by @dominsta.bsky.social !

Sora2 is speedrunning my AI law class. We covered issues with copyrighted characters in week 2, and right of publicity claims in week 3. Georgia has a postmortem right of publicity claim. Some states don't (e.g., famous Marilyn Monroe estate battle).

How Gemini Compute Use Agent feels about the "Choose Chrome" popup. gemini.browserbase.com

Why might AI companies take on larger copyright litigation risks? If they estimate AGI-scale impacts are 2-3 yrs out, litigation will lag that long. By then, the bet might be: govts step in (too big to fail), rightsholders reliant on AI, fair use prevails, or have $$$ to settle.

Quick take: Are open-weight AI models getting a fair shake in evals? A few thoughts on comparing systems-to-models, sparked by Anthropic’s recent postmortem. Check it our most recent post: www.ailawpolicy.com/p/quick-take...

GPT-5-codex just ``git reset --hard'' ongoing changes in a repo, saying "I panicked!" h/t Zeyu Shen @ Princeton

☢️ Can an AI model be "born secret" when it comes to nuclear and radiological risks? What powers does the Atomic Energy Act give the federal government over frontier models? It might be more than you think! And may preempt parts of state regs. Check out our post: www.ailawpolicy.com/p/ai-born-se...

Some quick thoughts on the recent copyright litigation developments: "Anthropic Settles Its Copyright Litigation—and Why That Was the Right Move" 🔗 www.ailawpolicy.com/p/anthropic-...

Annnnnndddd Judge Alsup just rejected the settlement. Still some time to fix it. Rejection was mostly on the grounds that the class was under-specified (no final list of works, no opt-out/notification mechanism solidified). news.bloomberglaw.com/ip-law/anthr...

💡New on the CITP Blog: "Statutory Construction & Interpretation for AI" > What if an LLM concludes a user's behavior is “egregiously immoral" -- & contacts authorities? CITP researchers with Prof @peterhenderson.bsky.social's POLARIS Lab provide a possible explanation.🔗👇

The terms of Anthropic's settlement w/book authors just came out. 💰$1.5B to authors in libgen (Books3 corpus)! Interestingly, this is ~$3k per book, close to the terms that HarperCollins allegedly gave to authors for their books ($2.5k). Consensus price forming?

Wonder why Claude decided to report users to the authorities? It might be because its constitution says Claude should choose responses in the long-term interest of humanity! But what if we could leverage computational and legal tools to "debug" or "lint" AI rules/laws for ambiguity? 🧵!

Excited to offer my AI Law class again @ Princeton this year. We'll be sharing lecture notes/materials and more this year on the course webpage! Imo, we have a unique offering that emphasizes how the technical details affect legal outcomes. Check it out! www.polarislab.org/ai-law-2025/...

I'm starting to get emails about PhDs for next year. I'm always looking for great people to join! For next year, I'm looking for people with a strong reinforcement learning, game theory, or strategic decision-making background...

A California teen sought advice from OpenAI's GPT-4o on how to end his life. The chatbot gave him explicit instructions and encouragement. His parents are suing the company and its CEO, Sam Altman, alleging “it was the predictable result of deliberate design choices."

Anthropic settled with authors in its ongoing litigation! Given the increasing likelihood of a messy trial, this was probably the best move. AI companies may have to be more strategic about which cases help set precedent in this area. Curious to see the terms.. news.bloomberglaw.com/class-action...

New paper suggests that if firms aren’t seeing growth from AI, it could be because current deployments replace existing labor, instead of scaling output. AI policy and governance agenda for 2025+ needs to put labor at the forefront. digitaleconomy.stanford.edu/publications...

Glad to see Google still working on efficiency-and transparency-of the energy impacts of their models!

AI-generated errors in an Australian murder case. We'll probably see an influx of ineffective assistance of counsel petitions/appeals soon arguing AI-usage. apnews.com/article/aust...

DOGE still exists, and continues to use Google's Gemini and other AI tools for identifying government regulations to slash. "The tool will further categorize [those] submitting comments, such as whether they [are a] 'sophisticated' corporate commenter." 👀 www.wired.com/story/sweetr...

New work from Hartline, Hu & Wu: is there a truthful calibration metric in sequential settings (i.e., better than ECE)? Seems like the answer is yes! Super important research direction as we think about multi-step uncertainty estimation from agents in high stakes settings.

Another judge potentially citing hallucinated case law? mississippitoday.org/2025/07/28/a...

Tesla partly liable in Florida Autopilot trial, jury awards $200M punitive damages

Check out our new blogpost and policy brief on our recently updated lab website! ❓Are we actually capturing the bubble of risk for cybersecurity evals? Not really! Adversaries can modify agents by a small amount and get massive gains.