Vaishnavi

I feel like I should attach the Michael Scott clip here about how I couldn’t possibly return to work immediately after vacation, but instead I offer this long weekend read of how a Renaissance diplomat made me think about child safety: quire.substack.com/p/long-weeke...

Just in time for this nor’easter: my early copy of @katieharbath.bsky.social’s new book is here! Katie’s view from the front lines of the biggest digital campaigns makes her my most important resource when thinking about how technology impacts politics and vice versa. Getting the tea (chai) ready ☕️

One positive about UNGA in town; you will always get your 10K steps without really thinking about it, purely from all the detours. This was at 10pm and the blockade was across multiple avenues and blocks so everyone got to know one another well!

Twenty countries now restrict children's access to social media. This fortnight: the EU KIDS Act, Meta's 332 CSAM, hundreds of Russian teenagers charged over online-recruited sabotage, and Kenya is building its evidence base.

I had the most emotional, raw reactions to @nhannahjones.bsky.social's reflections on education in New York City. I deeply respect the decision to live out her public values in private, yet have the courage to pivot when she realized her daughter was hurting from it.

Why must bike helmets be so ugly? Becoming familiar with NYC’s bike lanes but consistently constrained by the bulky black foreboding helmet I’ll have to carry around for the rest of the day in my travels. / rant

Made a few small changes to the AI Crisis Flowchart

AI-generated CSAM has been top of mind for companies since the Anderegg ruling a few weeks ago, so here's a short writeup on what's actually changed (and what hasn't) for child safety product roadmaps. quire.substack.com/p/ai-generat...

Yet another AI policy that nobody with product knowledge seems to have drafted. Who exactly is introducing purely companion chatbots into schools? The problem is educational AI tools that students inevitably turn to for emotional support, which of course this bill doesn't know how to address.

This was such a special podcast to be a part of with @yoyoel.com and @benwhitelaw.bsky.social. I hope that our stories can be valuable for other T&S professionals thinking about how to share their hard-won knowledge and expertise more widely.

In the ruling, the judge notes that Meta created a “public nuisance” by being a significant contributor to the state’s teen mental health crisis.

That's why I'm excited today to see @vaishnavi.bsky.social & team use Zentropi to build and release the first set of open source age-appropriate content policies to detect teen self-harm & suicidal ideation. Get them here: zentropi.ai/u/vys [3/n]

Making impossible safety challenges tractable is a lot of why we built Zentropi. Today @vaishnavi.bsky.social shipped a great example of what that means in practice - a set of open-source self-harm and suicide policies for teens: quire.substack.com/p/what-it-takes-for-llms-to-make-age 🧵 1/6

II members recently submitted input to NIST’s CAISI request for information on securing AI agent systems. Members highlight: ▶️ New risks from autonomous agents ▶️ Gaps in current cybersecurity ▶️ The need for stronger evaluation & safeguards Read the submission in full here: 🔗 bit.ly/4lUnMfy

If there's a better feeling than getting to inbox zero after a month on the road, I would very much like to know what it is. This is better than any spa.

Manhattan road closures for a half marathon are so maddening - so much traffic being diverted because a bunch of people needed a medal to get up and run this morning.

Losers: The Democrats, whom Americans blame for banning their favorite app; the Court, which spent its capital on a case it didn’t need to decide; the First Amendment, which the Court weakened; Congress, which gave us this dumb law; the American people, whose actual interests have been ignored. /1

Watching the Loomer / Musk showdown on the other site is bringing back fond memories of writing long, excruciatingly detailed policy assessments of whether online behaviours merit suspension. I’m glad that X continues to uphold its position on doxxing, whatever else might be going on there.

Deeply hoping I get to DC on time for FOSI, but a bridge strike in Baltimore may have other plans for me. PSA to rideshare drivers: get your 💰 by heading to an Amtrak station along the Northeast Corridor!

#trustandsafety friends this is such a great listen re: #safetybydesign. I loved @vaishnavi.bsky.social’s insights on SBD and safer online experiences for children and teens. Safe design benefits all, not just kids. m.youtube.com/watch?v=avkr...

A new issue of Quire in which we cover Australia's social media ban, the global push for age-based restrictions, and what fractured public trust in tech platforms means for them going forward quire.substack.com/p/social-med...

Great post for parents navigating their kids' use of social media by the always thoughtful @aagh.bsky.social

Feeling like a local - just clomped through a film set swearing mildly because they’ve blocked my street off and are obstructing my taxi (yes, taxi). Also, the muscle memory that just posts to X first is hard to correct. Tips welcome!

VYS’s next issue of Quire is out! While nuance around youth wellbeing and technology seems increasingly unlikely, here are some thoughtful ways to think about building responsibly for AI’s interaction with children: quire.substack.com/p/building-a...

I wrote for @techpolicypress.bsky.social about KOSA and what its popularity means for the future of child safety regulation in the US. Spoiler: Regulation is not always a net positive, but there are ways to make it work, which will require companies to rethink their approach to youth design.