iiinova — where your digital self keeps evolving

Reposted by Zander Arnao

Knight-Georgetown Institute (KGI) @knightgtown.bsky.social · 13/01/2026

🚨Why does access to public platform data matter? Join our webinar "Better Access: Data for the Common Good" (Jan 28, 2026, 11am-12pm ET) for a discussion on the Better Access framework, current regulatory shifts in the EU, UK + US, and what changes 2026 might hold. kgi.georgetown.edu/events/bette...

023

Reposted by Zander Arnao

Knight-Georgetown Institute (KGI) @knightgtown.bsky.social · 08/12/2025

🧵Our new model bill for US lawmakers showing how online platforms can be tasked with creating better algorithmic feeds was featured in Pluribus News. Read more here: pluribusnews.com/news-and-eve... /1

pluribusnews.com

Model bill offers blueprint for states to regulate algorithmic design – Pluribus News

112

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Didn't get to ask my question! But that's a wrap on #TSRConf. Really enjoyed attending this year and live skeeting. Thanks to @stanfordcyber.bsky.social. Y'all killed it!!!

010

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Meetali calls for more independent research on chatbots. For the Rain case (against OpenAI), TJLP benefited from more than 3200 pages of chatbot transcripts. This speaks to the power of data donations for fostering research

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

"We live in an environment where companies have gone from moving fast and breaking things to moving fast and breaking people." - @meetalijain.bsky.social Powerful words from a leading advocate in the field 🔥

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

David calls for academia to be more realistic. Trust and safety teams in companies are small and charged with many responsibilities. Academics could have more impact by studying solutions that do more with less

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Earlier this year - the judge in TJLP's case against Character AI ruled that it's unclear if the outputs of its chatbots are protected speech

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Challenges according to Meetali:the First Amendment and establishing that AI is a product. She calls for a statutory framework designating AI as a product to establish a cause of action. Open legal questions also exist - does a chatbot's output imply intent? Is intent necessary for accountability?

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Meetali on the law as a tool for promoting AI safety: while there's no dedicated state or federal chatbot laws, TJLP leverage product liability and consumer protection law (old and established doctrine) restricting unfair and deceptive practices

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

David from Meta distinguishes between "good" and "bad" engagement, arguing that engagement isn't a monolith. I'm going to try to ask him what he means by good and bad engagement during the Q&A

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Nate Fast: "Already by GPT-3, people preferred the interaction styles of chatbots over humans. It's a warning signal that people are attracted to these models. One of the concerns I have is artificial intimacy. It's easy to turn the dial up on this."

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

"I do believe litigation is the more important lever we have to effectuate change...I hope that we can put pressure and open up space from the outside which [other actors in the ecosystem] can leverage to create change." -- @meetalijain.bsky.social

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

@meetalijain.bsky.social rejects the term "companion." "It suggests friendship. These chatbots are not friends."

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

"I believe my role here is to issue an urgent warning call. We've never seen this kind of deluge of people who self-identify from being harmed by technology. These three cases are just the tip of the iceberg." - @meetalijain.bsky.social

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

@meetalijain.bsky.social starts her remarks with a story about Megan Garcia, whose son was sexually groomed by a chatbot. Meetali's org the Tech Justice Law Project brought three cases against leading AI companies: CharacterAI, Google, and OpenAI.

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Meta rep David Qorashi content that AI companions with empower users with great greater control over content and enable more transparency about content recommendations

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

I've been looking forward to this panel on AI companions with @meetalijain.bsky.social all day. This one is going to be spicy 🔥 #TSRConf

111

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Based on this analysis - children are to three types of harms - explicit, implicit, and unintentional. I'm a little unclear on the distinction between these three types of harms ❓

000

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

According to her research, harmful content is often framed as entertainment - eg offensive comedy or crime dramas - which can be problematic when exposed to children

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

And lastly: Haning Xue from the University of Utah on the role of algorithms in amplifying harmful content to children. Xue's study started with auditing the algorithm of Instagram, TikTok, and YouTube and the characteristics of content recommended to children

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Ofcom researches choice architecture using online randomized control trials to test small changes to safety features (eg increasing the prominence of user safety tools) and behavioral audits to systematically map design practices and evaluate their potential impact on user behavior

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Porter says design - the choice environment - matters because people are flawed decision-makers. Aspects of a platform can affect what consumers do. (Love the behavioral economics on display ❤️)

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Next up: Jonathan Porter from Ofcom (the British online safety regulator) on online safety! He starts with a spiel on the UK's Online Safety Act, which focuses in his telling on the backend of digital platform. Porter leads the UK's behavioral insights team and often examines platform design

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

CDT's recommendations: employers should assess the usefulness and necessity of hiring technology; deployments should adhere to accessibility guidelines (eg WCAG); and human oversight should be incorporated into all stages of using the technology

101

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Key findings: Workers with disability experienced a variety of barriers and reported feeling "extremely discriminated against." "They're consciously using these tests knowing that people with disabilities ren't going to do well on them, and are going to get screened out."

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Next up! The wonderful @arianaaboulafia.bsky.social at @cdt.org giving a talk on the exclusion of disabled workers by digitized hiring assessments. Background: companies are incorporating hiring technologies into employment decisions, which poses risks of discrimination and poor accessibility

111

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

The key finding: An overall increase in intimacy expressed by models over time. However, not all evaluation methods show a clear increase in intimacy over time

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

The research team evaluated 59 LLMs across 9 conpanies from 2018 to 2025 🤖 for the level of intimacy in expressed responses

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Next upu Pearl Vishen from UC Davis talk: "Is Intimacy the New Attention? An Audit of Expressed Intimacy Across LLM Generations" The key research questions: How does the level of expressed intimacy of LLMs evolve across generations? And has this gotten worse with subsequent generations of models?

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Lunch over! Now I'm attending a set of lightning talks on "AI and Algorithm Auditing" - starting with Rafal Kocielnik's talk on democratizing AI auditing #TSRConf

111

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

And that's it for the first part of the day! Lunch time 😋

010

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Good question from the audience: Is there any affordance for what happens when a platform shutters? This was a novel question for the framework authors. @naomishiffman.bsky.social imagines something like an Internet Archive-style data base that persists long-term

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

@leticiabode.bsky.social: "I think the strength of this framework is that it's more than just researchers. This data should be available for everyone." Hear hear!

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

A rep from YouTube: indemnification for platforms from future liability for sharing data in the manner they're obligated to would be another helpful incentive.

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Another interesting question: what's the pitch for companies on the value of sharing publicly available content? One thought from @naomishiffman.bsky.social: it's helpful to have a standard. It's been useful for companies in the past for companies to point to laws requiring transparency

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

@naomishiffman.bsky.social on this problem: should spotlighted content be made public? Yes. Many people have already seen that content, and it should be accessible to researchers and other external parties

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

A great point raised by an audience member: sometimes content because highly disseminated unintentionally. For instance, there have been cases of non-consensual intimate imagery being shared widely. Should the law require that content to be shared?

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Lengthy discussion about who is trustworthy and verifying trust from accounts who try to access data. Excellent example of how privacy poses challenges for greater transparency online 🙅‍♂️

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Group 3: How are data access mechanisms implemented? How is accuracy ensured, independence guaranteed, etc? A rep from TikTok asks a question around auditing: who is a trusted auditor? Interesting challenge! 🤔

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Group 2: What are the relevant thresholds for making content "publicly available"? Answer: the policy would only apply to big providers. Research on smaller services is less supported by this framework.

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Another complicated factor: European privacy laws. @daphnek.bsky.social says that content that identifies a person or is from an identifiable person is subject to data protections.

110

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

@naomishiffman.bsky.social and @leticiabode.bsky.social this data is fundamentally public and should be made public. Illegal content is more complicated. It should be shared, unless there are laws preventing someone storing the illegal content.

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Before I walked in everyone wrote down different questions on post it notes. It's taking a few minutes to categorize them into groups. Group 1: What if data is behind a login or obstructed by a platform's terms of service?

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Missed the beginning, but I walked in a discussion on how researchers at public universities can't scrape content online. But that creates tremendous legal risk! Case in point for the value of publicly available social media data 🔨

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

Now I'm heading over to @leticiabode.bsky.social's workshop!

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

* Long-term assessments - platforms conducts thousands of experiments per year. Engagement goes down in the short term with many design changes but are better in the long-term. If platforms held longer-term experiments, these impacts could be captured 📖

100

Zander Arnao @zanderarnao.bsky.social · 26/09/2025

* More user choices and better defaults - users should have at least one choice optimized for long-term user value. For minors, this feed should be the default 🙌

100