cute how the dumb betting market sites tried to pull a ‘monitoring the situation’ bar stunt and then NASA shows up and is like ‘pipe down pipsqueak, adults are working’

Doctors in Arizona, New Jersey, Ohio, Oklahoma, Texas, and Washington say that "the addition of AI-driven prior authorization means their patients are navigating a more complex and cumbersome system than they did a few months ago." wpintelligence.washingtonpost.com/topics/2026/...

their claim is thermodynamically inane, what are we even doing here

really worth a listen www.searchengine.show/the-trial-of...

continue to maintain that “AI policy” is a term that can mean too many things to too many people thereby making it a pretty unhelpful term. both under- and over-inclusive. do you work on consumer protection? export controls? labor policy? data centers? military use? ???

@afedercooper.bsky.social and I have a new paper that expands AI memorization tests to include near-verbatim copies, not just exact copies. We find that increases extraction/memorization of copyrighted content significantly. arxiv.org/abs/2603.24917

CVS 🚈 WAS

New paper on algs for clinical decisions and disparities: www.ahajournals.org/doi/full/10.... led by my wonderful student Safiya Sirota (soon on the market) TLDR: making decisions based purely on risk scores is not just suboptimal but may have consequences for racial disparities in treatment rates

in many ways i find myself feeling like we are back in 2015 or 2016 when it comes to understanding AI systems — of course so much has profoundly changed, but the feeling of reasoning through murky territory feels very familiar

Tasking an algorithm with making determinations about medical care can create unwarranted—and even discriminatory—delays or denials of necessary treatment. We're suing to get answers about Medicare's AI experiment. eff.org/press/relea...

Voters want jobs, not checks - "creating good-paying jobs" beats "direct income support" 54-17 as the preferred approach to AI displacement. And voters want a tax specifically on companies that profit from AI to pay for it (over a wealth tax) 49-27.

Excited to be on Odd Lots to talk about the politics of AI! AI today is less important than it will ever be. Over the past year, AI rose in issue importance faster than any issue we track — it's now more important to voters than climate change, child care, and abortion.

WAS 🚈 CVS

a three-page, bullet-points-only document with multiple competing policy proposals and no legislative text is under zero circumstances a credible policy framework.

Colorado's AI Policy Work Group is proposing a substantial re-work of the state's 2024 AI law. A few thoughts: andrewclearwater.substack.com/p/the-first-...

a fun side effect of learning how to ride a bike is you can see more planes

1/ As the US continues AI-enabled strikes on Iran, new @brennancenter.org research from @emileayoub.bsky.social and me examines how the military’s investments in the technology have been building to this moment. Congress must step up and reckon with AI’s dangers to life, liberty, and democracy 🧵:

more folks working on 'ai policy' — specifically policies seeking to prevent discrimination by ai systems in part through the use of impact assessments and audits — should read this paper by jennifer wang, @aselbst.bsky.social, @s010n.bsky.social, and @geomblog.bsky.social. arxiv.org/pdf/2506.17347

more folks working on 'ai policy' — specifically policies seeking to prevent discrimination by ai systems in part through the use of impact assessments and audits — should read this paper by jennifer wang, @aselbst.bsky.social, @s010n.bsky.social, and @geomblog.bsky.social. arxiv.org/pdf/2506.17347

Great resource for anyone engaged or interested in US AI policy and governance from @geomblog.bsky.social and his team www.brown.edu/news/2026-03...

NYT’s reporting on the DOD-Anthropic dispute sheds more light on how the Pentagon planned to use Claude for mass surveillance of Americans: by collecting and analyzing commercial bulk data. So what is this commercial data? And how might Claude be used to collect and analyze it? 🧵 1/

this piece does a good job of recapping this weekend's events www.transformernews.ai/p/openai-pen...

Ossoff’s statement is 1000x stronger than Schumer-Jeffries. This is a moderate Jewish Senator from a Trump state. The split isn’t moderate vs progressive, but leading vs poll-reacting

So, so instructive to compare media response to Afghanistan withdrawal -- months of hysterical garment-rending & catastrophizing, relentless criticism of the admin -- with the response to an unprovoked, unmotivated, illegal war of aggression, ie, "some critics say maybe Trump should have reasons."

"the Pentagon wanted the company to allow for the collection and analysis of unclassified, commercial bulk data on Americans, such as geolocation and web browsing data, people briefed on the negotiations said." www.nytimes.com/2026/03/01/t...

OpenAI claimed two things yesterday: 1. DOD "has no legal authority" to buy commercially available data (they have done this in the past and argue they do) 2. that "we can't protect against a government agency buying commercially available data sets." a bit contradictory

the US is now engaged in an illegal war against Iran. Congress has provided no authorization or declaration. given that: is the fight about DOD "all lawful purposes" language for AI companies really just downstream of questions about engaging in an illegal war?

i am curious how this is supposed to play out in the next few weeks: • OpenAI + Google may/may not have fights in signing up to replace Anthropic at DOD • if one lab tries to articulate a similar red line, DOD may well try to label them a supply chain risk • ??? www.defenseone.com/threats/2026...

while DoD threatening to use the Defense Production Act to force Anthropic to drop guardrails is a worrying escalation, "more important than what Anthropic has not agreed to is what other AI companies appear willing to accept [...] giving the DoD free rein to use their tools without restrictions."

it's officially been 11 years working @upturn.org. a few things to share from reflecting the past few days: • early on, friends and family didn't really understand my job. the intersection of technology and social justice? what? this almost seems silly now — it clicks for everyone immediately.

hello. a long-shot: if someone is willing to serve as an emergency FAccT reviewer for a paper re: red teaming and LLMs, please do let me know. I'm the AC for the paper and a reviewer has been unresponsive. thank you!

people continually dismiss how historically recent the enactment and enforcement of US civil rights laws are. conservatives will routinely dismiss discrimination as a thing of the past, not something that fundamentally continues to structure the present.

CVS 🚈 WAS

The ACLU (including my team) has filed multiple actions to block subpoenas sent by DHS to identify people speaking out against its abuses. Today's NYT story shows that those cases are the tip of the iceberg: DHS has sent HUNDREDS of subpoenas to tech companies. www.nytimes.com/2026/02/13/t...

New story up on the Homeland Security Department’s new tactic: flood social media companies with subpoenas to unmask anonymous accounts that criticize ICE or monitor the movements of ICE agents. www.nytimes.com/2026/02/13/t...

walk and chew gum, etc

Facebook plans to put facial recognition in its glasses and they think we’re too stupid to fight back. Their internal memo: “We will launch during a dynamic political environment where many civil society groups that we would expect to attack us would have their resources focused on other concerns.”

WAS 🚈 CVS

@reckless.bsky.social for your consideration for this week’s podcast within a podcast

hello! curious if anyone knows the following: what consulting firm specializing in civil rights law the Board of Governors of the Federal Reserve System contracted with in 1978 "to improve its efforts in assuring compliance with civil rights statutes." is the firm's study available publicly?

This is the second account - the first being in Chicago - of DHS agents mass detaining people and then *sorting them by race*.

I confess I do not understand why “the left should prepare for the worst case scenario with AI and jobs” seems to have elicited such a strong negative reaction bsky.app/profile/chri...

so for the DOJ, when a lender engages in predatory targeting of Hispanic borrowers, the way to settle the case is provide zero money to victims of discrimination, but instead have the lender spend $20 million on immigration enforcement in the areas in which the lend. www.justice.gov/opa/pr/civil...

PHL 🚅 WAS

WAS 🚅 PHL

This Equifax takedown letter from Senators Warren, Wyden, and Sanders is encouraging, but it's insane we need Senators to demand public price transparency from a contractor that works for the public

something proponents of algorithmic impact assessments should carefully assess is how new regulatory interventions (even in distinct technological contexts) with compliance burdens have or have not cashed out the past few decades.

DSA compliance can be massively wasteful. I testified about that to the EU Parliament during drafting. It's gotten worse since I wrote about it in 2024. It's really depressing for Trust & Safety people to spend their days on elaborate, pointless transparency reporting for an audience of zero. 23/

this is a wicked man who knows he is being wicked and does it anyway