@authorsalliance.bsky.social is doing great work and I was very pleased that they reposted my piece.

The Authors Guild v. OpenAI ruling just set a troubling low bar for "substantial similarity" of fiction summaries. If ChatGPT outputs are infringing, then thousands of Wikipedia plot summaries are next. matthewsag.com/copyright-wi... #CopyrightLaw #AI

A great roundup of anarchist briefs in Thomson Reuters' The Ross Intelligence. www.lawnext.com/2025/11/no-o...

Australia has so much electricity from solar power that it is going to start offering free electricity to everyone for at least three hours during the day as the wholesale price of power goes negative electrek.co/2025/11/04/a...

Harvey.AI is courting law schools like Westlaw/Lexis once did, but do we really need it? After a demo, my verdict: it's ChatGPT in a black turtleneck. Good for firms needing secure doc review & RAG, but law students can already do most of this with existing tools. matthewsag.com/do-law-schoo...

chatgptiseatingtheworld.com/2025/10/30/d... As predicted, but I think music generation is something of a special case.

Claude is showing some resistance to being tested on memorization. claude.ai/share/603385...

How should law schools adapt to generative AI? my new paper, AI Policies for Law Schools, proposes a template policy and explains why we should prepare students for legal practice as it will be, not as it was. 📄 papers.ssrn.com/sol3/papers....

Wispr Flow is amazing, I can't recommend it highly enough, but it can interpret random sounds with something like this: "For more information, visit www.fema.org"

My new post on competition from AI music and what it means matthewsag.com/competition-...

Why can't you copyright an AI-generated image that looks creative? My latest explores the "Skater Beagle" puzzle - how generative AI creates genuinely new works from latent space, not remix, but still lacks the human authorship copyright requires. matthewsag.com/skater-beagl...

matthewsag.com/drafting-law... This might be helpful.

Photojournalist Bill Biggart died when the second tower collapsed on him on 9/11, but the film he shot that day and his flash card were recovered from the debris and developed, along with his flash card. You can see his amazing photos of 9/11, the last photos he took while alive, here:

A new NBER working paper by Stella Jia and Abhishek Nagaraj makes some stunning claims about the effects of pirated book corpora on large-language-model (LLM) performance. I have my doubts. matthewsag.com/piracy-proxi... #AI #Copyright #books3

Why is everything about WordPress so ugly and awkward these days?

Invisible disabilities are all around you. www.huffpost.com/entry/dad-bu... "Advocating can be confusing and hard when your disability is invisible. ... People with disabilities, invisible or not, shouldn’t have to explain themselves. Nor should accommodations for them be met at the lowest bar."

Is the EU really the "rights-based" leader in AI regulation? In "The Illusory Normativity of Rights-Based AI Regulation," Yiyang Mei and I argue that this widely accepted assumption fundamentally misunderstands how EU AI regulation actually works. papers.ssrn.com/sol3/papers....

Stuck at O’Hare waiting hours for united to let us off the plane. I’m beginning to question my life choices.

I have all discussions of the president muted, so I had to hunt for this

GenAI has kept me so busy on copyright law issues that I have not done any empirical work for a while. Switching from Stata to python with Claude is a revelation! You still need to do the conceptual work, but its amazing.

It features pieces from @orlylobel.bsky.social, @ariezra.bsky.social, Aleksander Goranin, @matthewsag.bsky.social, Peter K. Yu, @jessicasilbey.bsky.social, @nicholson.bsky.social, and @saragerke.bsky.social!

Just checked my after visit summary from my health care provider. Was sad to see that I have lost 12 inches in height. Just as well they use infallible humans and not silly hallucinating AI tools.

It's really impressive how quickly journalists are able to digest a 40 page opinion, talk to half a dozen sources about their reactions and get something in print. No LLM is going to replace @knibbs.bsky.social.

Honestly would love to get these two judges on a podcast together to duke it out

Always nice to be cited followed by "But that can't be right."

A lot of people got sold on the idea that copyright was going to be the silver bullet that killed generative AI, often by people with no legal background whatsoever. That was never going to happen.

I just published Anthropic’s multi-billion dollar loss in Bartz v. Anthropic is really a win (for AI) medium.com/p/anthropics...

Has anyone else noticed that Gemini 2.5pro is writing in breathless prose as its default?

Carnegie Mellon’s Department of English is thrilled to announce an exciting new PhD program in Computational Cultural Studies. We will begin to accept applications this Fall with an inaugural cohort to begin in 2026. www.cmu.edu/dietrich/eng...

I've been waiting for video from this conference to be published so that I could urge everyone who cares about AI and copyright to watch @matthewsag.bsky.social give the most compelling explanation I've seen of how these two intimidating topics overlap. youtu.be/U1nP_1rXhk0...

Had an incredible time at the Fourth Annual AI Roundtable!! Fantastic papers and conversations--huge congrats to @matthewsag.bsky.social and Charlotte Tschider for organizing & @emorylaw.bsky.social for hosting. 🎉 🎉🎉

i’m starting to get a feel for o3 & o4-mini, and they are NOT as advertised — drop-in replacements for o1 & o3-mini they’re agents. if you use them as agents, they’re much *better*, but if you use them as word calculators, they’re far worse they’re a new thing

A very cool paper shows that you can use the RL loss to improve story generation by some clever setups on training on known texts (e.g. ground predictions versus a next chapter you know). RL starting to generalize already!

You can add a song to my Spotify playlist for the London Marathon. I am 98% of the way towards my fundraising target, but I only have 1:45 music on my playlist for an hour and 45 minutes. I’m going to need more! Donate any amount to add a song. 2025tcslondonmarathon.enthuse.com/pf/matthew-sag

Very excited for the Fourth Annual Legal Scholars Roundtable on Artificial Intelligence next week! matthewsag.com/emory-law-ai...

I'm running the London Marathon 3 weeks! You can add a song to my Spotify Marathon playlist by donating any amount to a very worthy cause (2025tcslondonmarathon.enthuse.com/pf/matthew-sag). I am 87% of the way to my fundraising goal, but I really want to exceed it and get to at least 3,000 GBP.

Volume 74, Issue 3 is live: scholarlycommons.law.emory.edu/elj/vol74/is... The issue features two articles: "Crisis Pricing" by John F. Duffy and Richard M. Hynes (of @uvalaw.bsky.social) and "Safety Whistleblowing" by Ben Cavataro.

Glad to see at least some of my works made it into LibGen, makes me wonder what the best way is to opt into AI training? #Fairuse The article by Alex Reisner in the Atlantic is worth reading: www.theatlantic.com/technology/a...

Finally may have figured out why LLMs rhyme so compulsively: instruction-tuning. Training an LLM to respond "helpfully" to user queries may push models into more "pleasing" aesthetic forms.

On March 14, 2025, I submitted my comments to the Office of Science and Technology Policy in relation to the “AI Action Plan”. You can read my comments at matthewsag.com/copyright-an...

www.linkedin.com/pulse/first-... great analysis of the Ross intelligence case

New issue of my newsletter: “Books, AI, and the Public Good: A New Grant” — A Mellon-funded project to develop an ethical, public-interest way to incorporate books into artificial intelligence

Feeling a little subtweeted here, maybe, lol. It’s ok; it took a while for Campbell to be accepted and understood. Bibas is allowed to misunderstand Warhol. But it is a misunderstanding. Competition was deadly to the first factor in Warhol b/c he was competing against Goldsmith *with her own work.*

matthewsag.com/thomson-reut... My initial thoughts on today’s summary judgment ruling in the Ross intelligence case.

Yet another case of hallucinated citations in court. storage.courtlistener.com/recap/gov.us...

Call for papers: Legal Scholars Roundtable on Artificial Intelligence 2025 (April 10-11, 2025). Please see announcement for details. docs.google.com/document/d/1... To apply to participate, comment, or present, please fill out the google form:( forms.gle/EyUghVd23tEB...). @emorylaw.bsky.social

I'll be speaking at the Country Radio Seminar 2025, sponsored by the Country Radio Broadcasters, in Nashville, TN, February 21, 2025.

Given DeepSeek's privacy policy (not much) and data storage (China) and the fact that it's open-weight, there is surely a market opportunity for someone to run a version in the US? You probably could not run it in the EU because to the AI Act? Or maybe you can because its open source? Thoughts?