Thanks to @samleecole.bsky.social and @404media.co for covering this important issue!

Yet another reason to visit Philly? I was sad when Temple's attempt to save UArts fell through, but fingers crossed that this merger with the Library Company will happen #libraries #specialcollections #rarebooks

Great overview of where we were, are, and the fight ahead! Thanks for the shout-out too! 🛟 @dataindex.us @prbdata.bsky.social @apduorg.bsky.social www.prb.org/news/researc...

Some interesting explorations happening here re: automating dataset changes and risk assessment...

We ❤️ our data librarian community!

🧵 Some publishers & news organizations are blocking the Wayback Machine from archiving journalism, stifling access to the public record. 📣 If you rely on the Wayback Machine to preserve news & history, tell them not to block web archiving. ✏️ Sign the open letter👇 www.savethearchive.com/NewsLeaders

I've seen some of these project presentations multiple times now...fine, clear-eyed use cases, but do they actually work??? Have tried the Wikidata MCP with a few LLMs now and it takes multiple tries to successfully complete very basic queries...

Excellent new research from @cdt.org @mluria.bsky.social analyzes the dark patterns applicable to AI chatbots, including general-purpose systems (ChatGPT, Gemini, Claude) and “companion” platforms (Replika, Character.AI): cdt.org/insights/dar...

Data Rescue Project is teaming up with the Williams Institute, ICPSR and MAP for an LGBTQ+ Data Rescue Hackathon on May 28th. This event will focus on federal datasets that include sexual orientation and gender identity (SOGI). Open to all and a great way to pre-game Pride Month!

🙌 "This commons gathers what indigenous communities are building, what researchers are documenting, and what history demands we remember — so the next generation of decisions about AI is shaped by those most affected by it."

⚠️Nearly all federal data users (93%) have been impacted by the changes of 2025. 54% say the impact on their work has been major. 📢A new SSRS EMERGE report gives these data users a voice — and the findings are urgent. Read the report: ssrs.com/emerge-initi...

Now's a great time to support LaVonnia Moore, a rural Georgia librarian who was fired because she approved a library display (set up by community members) that included a book featuring a transgender character. Her legal case enters the mediation process this week. 📚 www.gofundme.com/f/help-lavon...

"...what happens when institutions stop valuing the cumulative work of forming vocation and judgment — the kind of work that takes time?"

Thanks to @theguardian.com for highlighting the people involved in the effort to save and demonstrate the importance of federal public data. ❤️🛟 www.theguardian.com/us-news/2026...

These five librarians are running for state political offices across the country. Here's why they're running, and why librarians make good candidates. bookriot.com/librarians-r...

I really appreciate your role in this city @septaphilly.bsky.social, but what in the marketing rage-bait is this?! Please hire local copywriters and fix this jawn! #philly #septa

Beautiful essay..."Humans are incredibly good at adapting. When the environment is untenable, we move. When institutions and infrastructure fail, we come together in mutual aid. And through shared experience, memory, and continuity across geography, we remember that home is what we make of it."

In honor of ant season starting in my house: "What could we make with AI if scale wasn’t an objective? ...AI for Ants: A zine about Small AI” draws inspiration from the degrowth movement, climate justice, and low-tech to re-imagine AI that is locally-rooted and more attuned to our natural world." 🙌

Outline of how to approach the problem of AI extraction of open data/content through shared incentives and reciprocal benefit instead of direct monetization: “reciprocal contributions — improved metadata, processing tools, derived datasets — may align better with mission than licensing revenue.”

At a time when governments are willfully failing to preserve all their records, librarians and activists are using archival work to do more than document history. Guerrilla archives—whether digital or analog—are generally nonpartisan acts of preservation to serve the public good.

I don't know who needs to hear this, but app.radiooooo.com is probably one of the best things on the internet...tracks dripping in nostalgia, curated by actual humans, no algorithms. Taste is an interesting thing that...at least for now...still needs us

Had a great time in SF recently speaking @archive.org, meeting many folks that @datarescueproject.org has been working with. One outcome is this pledge committing to preserve and expand access to US government data and research. Please read, sign, share, and take action! publicinformationtrust.org

“Data is not enough” from OEDP’s Fieldnotes on AI Data & Tech Lead Cathy Richards focuses on the opportunity cost of the “lack of time to process information” refrain. ⚙️ 🔗Read/Listen to more here: open.substack.com/pub/openenvi... #Data #OpenData #AI #EnvironmentalData #EnvironmentalJustice

Is anyone really surprised that Claude code is a mess? Hilarious that Anthropic is issuing copyright takedown requests...

Today, the Archive published a Disappearing Data Chronology--a timeline tracking changes in access to federal information under Trump, including major data losses and restorations, legal challenges to information takedowns, and threats to archival collections. nsarchive.gwu.edu/special-exhi...

It’s important to remember that narcissism is in the DSM 5 and can be devastatingly detrimental to those affected…NOT some kind of innocuous productivity hack. Klein’s invocation of McLuhan ("we shape our tools, and then our tools shape us") re: AI sycophancy here is worth thinking about

Interesting thoughts here re: restricting AI/scraper access to archives/libraries conflicting with open knowledge/anti-surveillance ethos: "The remedy for misuse...lies downstream — in law, in professional norms, in community accountability — not in restricting access beforehand just in case."

New post: What's next for HIFLD? This is an exciting development for the data rescue community. Our partners at @publicenvirodata.bsky.social and Fulton Ring have used our rescued data to create HIFLD Next! The creators invite feedback! www.datarescueproject.org/hifld-next/

hrm...I don't think independence for arXiv necessarily=enshittification but I do wonder what the impetus for this is. Will be interesting to see how it plays out...hiring a library-world person for this role would be a good sign, but the size of that salary will definitely attract other types

Appreciating the call here for coordination as "essential to prevent duplication, ensure consistency, and close the widening gap between well–resourced and under–resourced organisations."🙌 zenodo.org/records/1893...

Reading over these rationales it becomes really clear how reckless, arbitrary yet totally hate driven the cancelling of these grants has been. The fact that ChatGPT was employed to do so is especially worrisome alongside OpenAI's recent defense/weaponry contracts

New report from UCLA Williams Institute School of Law: "Removal of Sexual Orientation and Gender Identity from Federal Data Collections, January 2025 to January 2026" More: https://myumi.ch/jJ8Qg

"How can librarians fulfill their professional and ethical obligation to defend intellectual freedom in the face of threats to their reputations, their jobs, and even their physical safety?"

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...

Friendly reminder that opposing both the oppressiveness of authoritarian regimes and this reckless war is not a contradiction on any front. Peace and safety to friends, family, and all living things in the Middle East🕊️ (video of a happy young goat at sunrise Jebel Shams, Oman from a recent visit)

For this week's #MemeMonday, don't be like Aragorn! If you're able to, please consider publishing your data in a reputable repository. We here at DRP deposit our rescued data in #DataLumos from @icpsr.bsky.social

"The “censorships” that the site purports to counter are not internet shutdowns or broad restrictions on content such as those in place in China and Iran, but European restrictions on hate speech and illegal content such as those in the Digital Services Act, or the UK’s Online Safety Act."

Preserving The Web Is Not The Problem. Losing It Is. Recent reporting by Nieman Lab describes how some major news organizations—including The Guardian, The New York Times, and Reddit—are limiting or blocking access to their content in the Internet Archive’s Wayback Machine. As stated in the…

In good news this week! ...hopefully these can be restored by summertime

"​In a time when historical information is actively being deleted, disappeared, or obfuscated, and competing visions of nationhood and citizenship are put forward, how do we critically engage with the archive as a contested site?"

The extractive nature of LLMs can be seen in their exploitation of not only open content, but also in their damage to open infrastructures.

Awful Ai Books is a new blog by a librarian, including helpful tips on how to spot them: "This Awful Ai Book is actually trying to kill you" awfulaibooks.wordpress.com

🙌 "Every day, millions of Americans rely on federal data—often without realizing it...Data for the People! is a new podcast from the Data Foundation that shines a light on this essential but often overlooked infrastructure."

I'm starting to see chatter around the new Wikimedia Enterprise deal that basically amounts to "AI bad=Wikimedia platforms bad now," so this point bears repeating...that this isn't about capitulation but the necessary steps to maintain the integrity of the infrastructure and demand attribution

Potentially good use case, but would like to see some more qualitative analysis. Even if there is just a small percentage of error, if those are conceptually or semantically egregious and the whole corpus still needs to be checked and corrected by humans, what really is the efficiency achieved here?

huge loss of a real one here😔 I was just last week re-reading her brilliant piece on "vocational awe"...something I find myself coming back to again and again: www.inthelibrarywiththeleadpipe.org/2018/vocatio...

On the weaponization of personal data: "If combined with the facial recognition and social media monitoring commonly deployed by the Department of Homeland Security, those reams of data would turbocharge ICE’s terror campaign in the short term and destroy American civil liberties in the long term."

Hello #DataRescuers! We're excited to start this AMA. We'll be answering questions over the next 2 hours. Please feel free to like, respond, or reskeet as we go.

New from 404 Media: here is the user guide for ELITE, the tool Palantir made for ICE. A map shows people on a map; ICE clicks them, brings up their dossier. ICE uses to find which neighborhoods to target. We obtained it, you can now read the document for yourself: www.404media.co/here-is-the-...