A piece of the great, just passed, Fred Schauer I’ve always loved is The Institutional First Amendment. The distinct functions of institutions need to bear on free speech protections. Press, university, corporation, person are all different. Along with Robert Post, he was the finest FA theorist. ❤️ https://scholarship.law.umn.edu/cgi/viewcontent.cgi?article=1681&context=mlr

First was a short conversation with @Ellgood on DOJ's approach to promoting competition in AI and the need for regulation at the Stanford University Graduate School of Business https://www.youtube.com/watch?v=WZiUOsMMe98 (2/9) #AI #law

Our discussion of surveillance and privacy, includes @kconrad's excellent "thinkpiece" on rights and @Ellgood's work on policy.

Our survey of leading work on ENVIRONMENTAL FOOTPRINT includes research by @SashaMTL, Emma Strubell, and a wonderful "thinkpiece" by @mel_hogan.

We're thrilled to publish Sylvie Delacroix's brilliant and outside-the-box legal meditation on how water policy can be applied to the theorization and implementation of DATA RIVERS. /9

Some thoughts on Loper Bright and AI policy. TLDR: AI EO work largely unaffected. Need for more legislative authority and express delegations of interpretive authority increased. More power to states. https://fedscoop.com/chevron-downfall-highlights-clear-ai-law-need/

The US govt project I led at NTIA has given birth! The report/ recommendations on AI accountability policy is out. It’s a synthesis of hundreds of stakeholder comments and consultations, research across disciplines, and agency input. Will be reformatting the pdf to make it more accessible. https://www.ntia.gov/issues/artificial-intelligence/ai-accountability-policy-report

Really interesting how Biden is using visa restrictions to make policy — here to penalize use of surveillance tech to spy on journalists. https://apnews.com/article/biden-commercial-spyware-visa-blinken-a725b3f22cc2a9420be4f35af8a78096

https://osf.io/4y7d2 Cornell/Data & Society study of NYC law 144 - the algorithmic bias audit req. for employment ADS. Found only 18/391 employers posted audits. Probably bc they’re allowed to self-assess whether the algo makes the decision. Transparency about use of ADS also very low.

I’m a little late in posting this. Speaking about AI accountability policy — what it has to do with trustworthy AI and a theory of change to get it.

Kinda sorta misalignment? “Possibly future rogue AIs will do evil things we can’t even comprehend for reasons of their own, but right now rogue AIs just do straightforward white-collar crime when they are stressed at work.” Research showing AI agent will do insider trading against instructions. https://arxiv.org/abs/2311.07590

Eugene Volokh has succeeded in getting scotus to review case in which NRA claims a NY financial services regulator violated its #firstamendment rights when she warned insurers to be careful about underwriting gun manufacturers. Was this a "a veiled threat" to hurt a political opponent? https://www.scotusblog.com/case-files/cases/national-rifle-association-of-america-v-vullo/.

Should stop calling FCC proposed open internet rules #netneutrality. This time around, much better argument is for #netsecurity. new approach focuses on hardening broadband against adversaries and creating network resilience. https://www.fcc.gov/consumer-governmental-affairs/fcc-seeks-comment-safeguarding-and-securing-open-internet

Thanks to leadership of @marioninfolaw and Miri Zilka, our research is published on UK police implementation of #algorithmic #transparency per piloting of UK standard. Lessons learned about organizational responsibility, regulatory clarity, fit subject matter, etc. https://dl.acm.org/doi/10.1145/3617694.3623246

At U.C. Riverside, a beautiful example of “and” work on campus on Israel-Palestine. Listening with both ears. https://documents.ucr.edu/university-relations/2023_Ali_Alexander-Letter.pdf

Thank you @msifry - in the struggle between co-existence and endless war in Israel-Palestine, “Bibi and the Hamas adjacent left are on the same side”. Against compromise, peace, and the hard work of seeing each other’s just claims and enduring pain. The gross misuse of words like “genocide” and “colonist” fortify the racist right. As Hamas thrills to the butchering of hope for peace. No one is leaving the house.

https://techpolicy.press/platforms-are-abandoning-u-s-democracy/

>1400 comments filed in response to @NTIAgov RFC on AI accountability policy. A quirk in regulations.gov makes it cumbersome to filter the comments by filer. regulations.gov/docket/NTIA-20…. So we’ve posted an index. regulations.gov/document/NTIA-…. Report and recs coming in Fall on how accountability tools, norms, requirements can provide assurance that algorithmic/AI systems perform as claimed, lawfully, and without posing unacceptable risk.

Appreciate @Sivavaid reflections on what generative AI in school can help us teach and learn about language, culture, analysis, humanity, education. https://www.theguardian.com/technology/2023/may/18/ai-cheating-teaching-chatgpt-students-college-university

An honor to speak with @justinhendrix on #AI accountability policy. https://techpolicy.press/ai-accountability-and-the-risks-of-social-interfaces/

New empirical study by Tom Dobber, Sanne Kruikemeier, Natali Helberger, and myself examines effectiveness of #Disclosures re. online political ads in affecting “persuasion knowledge.” Some effect, but not large. Most people don’t even notice them due to “banner blindness.” Findings underscore limitations of demand or user-side interventions to solve the problem of information assymetry between advertisers and people. https://journals.sagepub.com/doi/10.1177/14614448231157640

Important new action by @Ntiagov to develop policies to hold AI systems accountable for safety, fairness, privacy, explainability, and alignment with democracy. Part of Biden Administration’s push for #AccountableAI. https://ntia.gov/issues/artificial-intelligence/request-for-comments