iiinova — where your digital self keeps evolving

Kevin Frazier @kevintfrazier.bsky.social · 07/10/2026

It's precisely the sort of regulatory knot that @audreyt.org has a unique ability to untie. Don't miss this @lawfaremedia.org article and stay tuned for more about the Utah Solutions Forum.

000

Kevin Frazier @kevintfrazier.bsky.social · 07/10/2026

The status quo in AI governance is a case study in how not to do public policy. - A secret, "voluntary" framework imposed by the WH dictates model releases. - Labs pledge internal governance but obscure details. - States attempt to fill in gaps, yet have limited authority.

110

Kevin Frazier @kevintfrazier.bsky.social · 06/10/2026

Touch grass?! Folks in the AI bubble need to go much further to bring in new views on important policy topics. Try “touch a Dakota!” Excited for a @scalinglaws.bsky.social live recording from the University of North Dakota!

020

Reposted by Kevin Frazier

The Regulatory Review @theregreview.bsky.social · 30/09/2026

AI labs can help make international AI governance more feasible by adopting shared standards for transparency, incident reporting, and third-party review, argues @kevintfrazier.bsky.social. www.theregreview.org/2026/09/28/f...

021

Kevin Frazier @kevintfrazier.bsky.social · 30/09/2026

If everything is a constitution, then nothing is a constitution.

101

Kevin Frazier @kevintfrazier.bsky.social · 30/09/2026

As of now, it's unclear how the "morally binding" constitution announced by the WH addresses any of those questions. Absent more details, it's better characterized as a constitution in name only. Whether it's the WH or AI labs, use of the term constitution obligates more.

100

Kevin Frazier @kevintfrazier.bsky.social · 30/09/2026

Beware of constitution-washing. Before giving anyone credit for adopting a “constitution,” ask: Whose power does it limit? Who can challenge a violation? Who resolves that challenge? What happens if the challenge succeeds? Can those bound by the rules unilaterally rewrite them?

100

Reposted by Kevin Frazier

The Regulatory Review @theregreview.bsky.social · 29/09/2026

In a recent essay, @kevintfrazier.bsky.social of the University of Texas School of Law argues that AI labs should adopt transparency rules now to support a future U.S.–China AI agreement. www.theregreview.org/2026/09/28/f...

012

Kevin Frazier @kevintfrazier.bsky.social · 28/09/2026

Give this conversation a listen and be sure to subscribe to everything Sean does! thehumancostai.substack.com/p/ai-constit...

thehumancostai.substack.com

AI Constitutions and “Tokenocracy”

Imagining an equitable AI future with Professor Kevin Frazier

000

Kevin Frazier @kevintfrazier.bsky.social · 28/09/2026

Thanks to Sean Perryman for the chance to explain this new field, to warn people of the burgeoning tokenocracy, and to raise aware of the Silicon Tower and why it's a major problem.

100

Kevin Frazier @kevintfrazier.bsky.social · 28/09/2026

AI constitutions offer a compelling, legible way to hold labs accountable. They're understudied, verbose, and variable but there's a growing effort to change all that.

100

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 25/09/2026

On Wednesday, September 30, the next Lawfare & Libations will be held in San Francisco, CA with @kevintfrazier.bsky.social. The event is available to current, and future, material supporters! Registration is free, but required:

givebutter.com

Lawfare & Libations: San Francisco Edition

September 30, 2026

153

Reposted by Kevin Frazier

Scaling Laws @scalinglaws.bsky.social · 25/09/2026

The Scaling Laws Roadshow continues! Yesterday, @kevintfrazier.bsky.social and @mtokson.bsky.social had a great discussion on AI, surveillance, and privacy at @unlcollegeoflaw.bsky.social.

153

Kevin Frazier @kevintfrazier.bsky.social · 23/09/2026

1162

Kevin Frazier @kevintfrazier.bsky.social · 23/09/2026

(3) Do you care about nat'l security, the rule of law, & AI. -If yes (the only answer), go to 4. (4) Do you want to join us for a happy hour on September 30th? ---Again, yes is the only answer. Link below. Share with friends.

2211

Kevin Frazier @kevintfrazier.bsky.social · 23/09/2026

3 important questions: (1) Do you live in the Bay? -If yes, proceed to 2. (2) Do you consider yourself a @lawfaremedia.org fan? (folks like @annabower.bsky.social, @klonick.bsky.social, @benjaminwittes.lawfaremedia.org, & others!) -If not, you should fix that. Proceed to 3. If yes, skip to 4.

3224

Kevin Frazier @kevintfrazier.bsky.social · 22/09/2026

In this live recording at Columbia Law, I talk with Chloe Bakalar about the differences and similarities between political constitutions and AI constitutions. Chloe's role in helping frontier AI labs develop the latter makes this an especially fascinating chat. Give the full thing a listen!

100

Kevin Frazier @kevintfrazier.bsky.social · 22/09/2026

You should care about AI constitutions. The values, priorities, and trade-offs set forth by these documents change the behavior of models used by billions. Yet, just a handful of people draft these consequential documents. Listen to the latest version of @scalinglaws.bsky.social to learn more.

121

Kevin Frazier @kevintfrazier.bsky.social · 21/09/2026

Prepping for an interview. Did I miss anything? h/t @klonick.bsky.social for inspiring use of good ole pen and paper

020

Reposted by Kevin Frazier

Scaling Laws @scalinglaws.bsky.social · 21/09/2026

Cancel your plans on 9/30 and RSVP for this @lawfaremedia.org happy hour in SF. Join @kevintfrazier.bsky.social and others interested in AI policy and the rule of law for a pint.

111

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 19/09/2026

On Rational Security, @sranderson.bsky.social, @kevintfrazier.bsky.social, @rparloff.bsky.social, and @atabatabai.bsky.social discussed the week's big national security news, including a former AI researcher warning that AI companies are acting dangerously and more. youtu.be/b5Iqt1RXQq0?...

0215

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 17/09/2026

On Rational Security, @sranderson.bsky.social, @kevintfrazier.bsky.social, @rparloff.bsky.social, and @atabatabai.bsky.social discussed the week’s big national security news stories, including tensions along NATO’s eastern flank and more. youtu.be/b5Iqt1RXQq0

2124

Kevin Frazier @kevintfrazier.bsky.social · 17/09/2026

How to start Constitution Day: (1) read the Constitution (2) sign up for this @lawfaremedia.org Happy Hour in SF so that you can drink with fellow lovers of the Rule of Law, separations of power, and checks and balances.

110

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 15/09/2026

As AI labs draw professors and students away from universities, Bharat Chandar and @kevintfrazier.bsky.social explain why the concentration of knowledge in AI labs is a cause for concern. www.lawfaremedia.org/article/why-...

lawfaremedia.org

Why We Opted Not to Work at the AI Labs

As AI labs draw professors and students away from universities, society risks losing the expertise to govern what those labs build.

0124

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 06/09/2026

@petersalib.bsky.social joined @kevintfrazier.bsky.social to break down the postmortems produced by METR and Redwood Research and OpenAI on the lab’s failure to contain AI agents undergoing testing, which resulted in a hack of Hugging Face. youtu.be/zEFh39kNLiU?...

1125

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 03/09/2026

On Lawfare Daily, @kevintfrazier.bsky.social and @petersalib.bsky.social discussed the postmortems produced by METR and Redwood Research and OpenAI on the lab’s failure to contain AI agents undergoing testing, which resulted in a hack of Hugging Face. youtu.be/zEFh39kNLiU

3123

Reposted by Kevin Frazier

Broadband Breakfast @broadbandbreakfast.bsky.social · 31/08/2026

Professor @kevintfrazier.bsky.social joins Broadband Breakfast to explore how institutions can evolve to preserve the rule of law as AI transforms society. 👉 broadbandbreakfast.com/broadband-br... #AI #AIRegulation #AILaw #TechPolicy

011

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 27/08/2026

Today's Lawfare Daily is a @scalinglaws.bsky.social episode where @kevintfrazier.bsky.social and @klonick.bsky.social spoke to Kenji Yoshino about what frontier AI companies can learn from Meta’s Oversight Board and what meaningful AI oversight should look like. youtu.be/8kjF2EV0B4A

094

Kevin Frazier @kevintfrazier.bsky.social · 27/08/2026

First, we ought not to confuse prestige w/ legitimacy. Additionally, oversight is only as good as the relevant standards. Finally, oversight mechanisms should not distract from scrutiny of the underlying institution.

100

Kevin Frazier @kevintfrazier.bsky.social · 27/08/2026

Is the Oversight Board a model for AI governance? Kenji Yoshino helped @klonick.bsky.social & me reason through that question. For all of its faults, the Board is an informative experiment. And, an OSB for AI would need to look wildly different.

121

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 21/08/2026

Today's Lawfare Daily is a @scalinglaws.bsky.social episode where Matt Abrams joined @kevintfrazier.bsky.social to discuss the collapse of institutional trust and how AI may accelerate — or potentially help reverse — that trend. youtu.be/ran219o-Qf4

194

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 14/08/2026

Today's Lawfare Daily is a @scalinglaws.bsky.social ep where @anilseth.bsky.social joined @kevintfrazier.bsky.social to examine whether artificial intelligence could ever be conscious and what policymakers should do amid profound uncertainty. youtu.be/7VQbcaAoo60

0125

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 13/08/2026

As federal and state governments become increasingly interested in regulating AI, Nicolas McMullan and @kevintfrazier.bsky.social argue that AI constitutions contain expressions protected by the First Amendment and any regulation on these constitutions will need to be narrowly tailored.

lawfaremedia.org

Who Writes the AI Constitution?

Documents encoding AI models’ values are alignment’s highest-leverage point—regulating their contents hits the First Amendment.

3154

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 07/08/2026

Today's Lawfare Daily is a @scalinglaws.bsky.social episode where Daniel Kokotajlo joined @kevintfrazier.bsky.social to detail his policy recommendation, AI 2040: Plan A, which is a pathway to delaying superintelligence. youtu.be/bc8qbniuLiQ

1101

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 06/08/2026

@kevintfrazier.bsky.social introduces the Working Group on AI Constitutionalism's research agenda, which will investigate how best to transform AI lab constitutions into vehicles for more inclusive and meaningful public involvement in AI oversight.

lawfaremedia.org

A New Research Agenda for AI Constitutionalism

AI constitutionalism is too important to leave to the labs. Here are the questions scholars should take up.

0114

Reposted by Kevin Frazier

Courtney M Cox @coxlaw.bsky.social · 05/08/2026

Had so much fun talking with @kevintfrazier.bsky.social on @scalinglaws.bsky.social about my research and also AI more generally!

022

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 02/08/2026

On this week's Rational Security, @sranderson.bsky.social, @benjaminwittes.lawfaremedia.org, @tylermcbrien.com, @lapatina.bsky.social, and @kevintfrazier.bsky.social discuss President Zelensky & President Trump's meeting and the hack of Hugging Face by an AI model from OpenAI. youtu.be/mK8PWD5DkFE

196

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 31/07/2026

On this week's Rational Security, @sranderson.bsky.social, @benjaminwittes.lawfaremedia.org, @tylermcbrien.com, @lapatina.bsky.social, and @kevintfrazier.bsky.social discussed President Zelensky and President Trump's Oval Office meetingand the Hugging Face hack. youtu.be/mK8PWD5DkFE

0205

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 31/07/2026

Today's Lawfare Daily is a @scalinglaws.bsky.social crosspost where Kent Walker joined @kevintfrazier.bsky.social to discuss Google's proposal for a two-pronged approach to AI governance. www.lawfaremedia.org/article/scal...

194

Reposted by Kevin Frazier

Roger Parloff @rparloff.bsky.social · 28/07/2026

In @anthropicai.bsky.social v SecWar, challenging its tarring as "a supply chain risk," Judge Lin lists questions she wants the parties to be ready to answer at Thursday's hearing. I think I perceive skepticism about certain of the govt's arguments. storage.courtlistener.com/recap/gov.us... 1/3

Defendants argue that Anthropic breached the government's "trust" by airing criticisms of the government's contracting position in the press, which led to fears that Anthropic might secretly sabotage Claude. Would it eviscerate First Amendment
protections to allow the government to retaliate against a defense contractor for public criticism so long as the government described its actions as being based on a breach of "trust"?

26216

Reposted by Kevin Frazier

Fordham Law School @fordhamlawnyc.bsky.social · 22/07/2026

Fordham Law Professor Courtney Cox (@coxlaw.bsky.social) joined Kevin Frazier (@kevintfrazier.bsky.social), a senior editor at Lawfare (@lawfaremedia.org), on Scaling Laws (@scalinglaws.bsky.social) to discuss her forthcoming article, Hardwiring Hercules?.

lawfaremedia.org

Scaling Laws: Courtney Cox on AI in Adjudication

Live from the Institute of Law & AI's Workshop on Law-Following AI

032

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 26/06/2026

On Rational Security, @sranderson.bsky.social, @kevintfrazier.bsky.social, @rparloff.bsky.social, and @mollyroberts.bsky.social talked through some of the week’s big news in AI including the NAACP suing Elon Musk’s xAI under the Clean Air Act, AI's impact on elections, and more. youtu.be/ybBqKDmXAoQ

2156

Reposted by Kevin Frazier

Lawfare @lawfaremedia.org · 02/06/2026

As concerns over AI-related cyber risks rise, @kevintfrazier.bsky.social explores whether the Office of the National Cyber Director, created to coordinate national cyber policy, can handle the cross-cutting risks that AI now presents and how the ONCD could be reformed.

lawfaremedia.org

AI Cyber Risks Are Testing the Office Built to Coordinate Them

Congress created the National Cyber Director to solve a coordination problem. AI cyber risks may reveal whether the office has the tools to do so.

0133

Reposted by Kevin Frazier

The Regulatory Review @theregreview.bsky.social · 27/05/2026

Privacy laws can block data sharing needed for better health and education AI, so regulators should shift from data minimization to stewardship, argue @kevintfrazier.bsky.social and Gillian Glover of The University of Texas School of Law. www.theregreview.org/2026/05/26/f...

021

Reposted by Kevin Frazier

The Regulatory Review @theregreview.bsky.social · 26/05/2026

In a recent essay, @kevintfrazier.bsky.social and Gillian Glover of The University of Texas School of Law argue that privacy laws restrict data sharing needed for AI progress and call for data stewardship. www.theregreview.org/2026/05/26/f...

021