Chapter 3 is up!

one of the most important conversations tech companies should have about every new feature and product is "how can this be abused" and in OpenAI's case they apparently decided to make that the selling point for their new app this is a million lawsuits and ruined lives waiting to happen

Find someone to hate the way Eliot Cohen hates Pete Hegseth. OMFG www.theatlantic.com/ideas/archiv...

I’ve had experience with a lot more generals and admirals than the average left-leaning critic of the military-industrial complex and 21st century wars. There’s an occasional jerk or narcissist, but generally they are thoughtful, committed, allergic to bullshit. Hope my sample is representative.

Waymo dodges a ticket b/c tickets lack a field for robot drivers. Imagine if it got points for a moving violation, and every cars' points went to one AI's license, those could add up pretty quickly. www.theguardian.com/technology/2...

Been looking forward to reading this for a while!

"The Internet Archive isn't scraped by AI—they prohibit the bots from training on their massive trove of data. So from the perspective of AI, the Internet Archive simply does not exist."

I like this -- they fed AITA posts where the humans voted "asshole" to LLMs and found that the bots would almost always tell you that you're not an asshole. ^^; No wonder they're so popular! www.businessinsider.com/reddit-aita-...

🚨 New Publication 🚨 ISP Fellows Rebecca Crootof & @bjard.bsky.social have launched their Technology Law coursebook! First chapters now live: 📑 TOC: papers.ssrn.com/sol3/papers.... 📘 Ch.1: papers.ssrn.com/sol3/papers.... 📘 Ch.2: papers.ssrn.com/sol3/papers.... New chapters drop twice monthly.

Nightstand Pile Files: Re-reading @jackbalkin.bsky.social's "Old School / New-School Speech Regulation," on governments' various techniques for controlling and surveilling speech. I read it nearly every year. And every year it feels more and more relevant. papers.ssrn.com/sol3/papers....

Chapter 2 is up!

It has been 10 days now since the Trump administration conducted a lethal strike on a boat in the Caribbean. In that time, the administration has not even attempted to present a legal justification for this premeditated killing of people outside of armed conflict.

🧵 The summer of 2025 has been AI's "cruel summer"—wrongful deaths, dangerous therapy chatbots, medical misinformation, facial recognition failures. These aren't isolated glitches but predictable harms from systems deployed without adequate oversight. www.science.org/doi/10.1126/...

Nightstand Pile Files: Re-reading @rcalo.bsky.social's excellent "The Scale and the Reactor," on what techlaw and STS scholars can learn from each other. Aside from informative and prescriptive, it is jam packed with delightful phrasings and analogies. papers.ssrn.com/sol3/papers....

My school doesn’t offer a degree in Outlook, even though odds are very good that a majority of my students will be required to use it in their professional lives, and many of them will use it poorly/incorrectly.

We’ll release new chapters biweekly with updates to our front matter and table of contents at docs.google.com/document/d/1.... Chapter 1 introduces our "techlaw uncertainties" approach at docs.google.com/document/d/1.... We welcome your feedback!

Second, I’m thrilled to share the first excerpts of the Technology Law coursebook I’m co-authoring with @crootof.bsky.social. We’ve built these materials across courses at three institutions, beginning with the seminar we taught with @jackbalkin.bsky.social at Yale Law School in 2016.

The United States has committed every moral atrocity that has ever been committed. I can’t prevent that. What I can do is say that it should not do so, and if it does so, we should call it what it is, condemn it, hold ppl accountable for it, and deprive them of power.

As many know, @bjard.bsky.social and I have been drafting a Technology Law coursebook for a few years. We've used it to teach classes at three institutions, including Yale Law School, and others have used chapters in their techlaw classes. We're excited to share the current version more broadly!

I'll admit, I was skeptical when they said Gemini was just like a bunch of PhDs. But I gotta admit they nailed it.

syllabus time, teaming up for the herculean efforts of reinventing writing assignments. here, some prompts for required weekly low stakes 250-500 word reflections/ process pieces that have proven relatively conducive to real writing in lit class. please share any similar suggestions in thread.

i keep coming back to this.

Anyone working on inserting secret "sleeper" viruses in their novels that can be activated only inside LLMs? Perhaps along the lines of Snow Crash.

Current measures of #CivilianHarm do not adequately account for the aggregate impacts of protracted armed conflict on civilian populations, exemplified by the war in #Gaza. Fionnuala Ni Aolain offers the concept of cumulative civilian harm as an analytic tool: www.justsecurity.org/115407/cumul...

Tech companies only have five ideas: robot slave (actually just human slaves), hallucinatory counter-reality, untaxable money, The Everything App, and Clippy

everyone in the natsec and FP community increasingly just losing their minds as the topical questions in US FP and natsec these days increasingly devolve to the level of "should we lick this toaster while it is on? some experts weigh in"

A huge step backwards. To save nearly no money, in military terms, we incur such a deep cost.

Nightstand Pile Files: Military AI as Sociotechnical Systems by Mitt Regan Mitt Regan talking about how miliary AI decision assistants should be regulated as sociotechnical systems. What's not to love??? A great intro for those new to any of these issues or teaching an "AI & Law" class.

19/18: @haaretzcom.bsky.social archived here without paywall. “Enough is Enough. Israel is committing war crimes” by ex-Likud conservative former Prime Minister Ehud Olmert: archive.is/2025.05.28-1...

Nightstand Pile Files: Chain-Gang All-Stars by Nana Kwame Adjei-Brenyah Wow. Hunger Games meets systemic racism/the carceral state/abolitionism . . . with FOOTNOTES. With a dash of unintended side effects of well-intentioned research. And distinctive, understandable, deeply human characters.

Stanislav Petrov, the man who saved the world, died on this day in 2017. If you don’t know the story, just read this www.theguardian.com/world/2017/s...

Nightstand Pile Files: Transnational Atrocity Victim Networking: A New Paradigm for International Accountability Mechanisms @profgsgordon.bsky.social A creative proposal for connecting and empowering victims of wartime harms, to facilitate participation and compensation. Implementing #WarTorts

Red teaming is great for identifying one source of risk - namely, malicious acts and the accident risks that they foreshadow. But absent intentional consideration of likely harms that arise when everything works WELL and tech proliferates, red teaming won't anticipate all likely risks and impacts.

Huge !!

Truth.

Nightstand Pile Files: The U.S. DoD's New CHMR Policy and "Harm Resulting From" Military Operations @danmaurerlawprof.bsky.social The DoD definition of "civilian harm" has the qualifier "resulting from military operations" - what causation test should be used to evaluate that? #WarTorts

the fundamental disconnect between the manosphere and reality is that they are incapable of understanding that this is - according to every woman I’ve ever met - the sexiest thing a man has ever done

The thought experiment was right: we have invented a paperclip machine.

Truly impressive investigative work. Indeed difficult, but vital to watch. In addition to @janinadill.bsky.social's analysis, a few LOAC notes: --no claim of warning prior to firing on ambulances --commander's failure to punish can also be war crime --IDF personnel may face universal jx abroad

it's useful to think that all science had to be funded into existence and prior to its modern professionalization, almost all science was carried out by the idle rich (bros) that is, the science that exists is the science that people wanted to exist, and a lot of the time it's to argue something

Nightstand Pile Files: AI Red-Teaming is Not a One-Stop Solution to AI Harms @friedler.net Ranji Singh @borhane.bsky.social Jacob Metcalf Brian J. Chen Admittedly, I've been too often defaulting to "Wouldn't red teaming address that?" This is a concise clarifier of when (and how) it can - or won't.

Fearing that Meta’s chatbots were too boring, Mark Zuckerberg ordered they be given more human-like traits. That rebuke led to a loosening of boundaries -- and the launch of digital companions that Meta knew could produce sex role-play content for children. www.wsj.com/tech/ai/meta...

Nation Can’t Believe It On Harvard’s Side

I’m sorry but if your students use AI to write papers and you use AI to grade them zero school is happening. You are running together on a hamster wheel

Even accepting the premise that AI produces useful writing (which no one should), using AI in education is like using a forklift at the gym. The weights do not actually need to be moved from place to place. That is not the work. The work is what happens within you.

"[Australia's] recognition of the need to compensate victims is welcome. [...H]owever, Australia still regards compensating victims of war crimes as its charitable gift, rather than as an international obligation...." lieber.westpoint.edu/australian-c... #WarTorts

This whole thread.

May need to start including this as boilerplate in the star footnote.

Two of my articles. Not sure if it's for better or worse that they're not my best ones? www.theatlantic.com/technology/a...

I want to make the below thread more concrete and, in the process, to accentuate what I take to be the enormous challenge before us. So, here’s a second thread that starts by drawing on my work with Ingrid Brunk on annexations @ajil.bsky.social. 1/ 🧵