I’m working on an AI policy for my org that allows us to opt out of AI note taking and prohibits AI in our comms/storytelling. here is my list of reasons for the policy, but my board is asking me to cite sources. Can you help me with any good references you would cite for any of these? (Or an edit or restatement where I’ve gotten it wrong or inaccurate?)
*if you want to argue about why I shouldn’t have this policy kindly crawl into a hole in the ground and cover yourself with soil
@aud @seachanger That's about the only actual study we have and it has a fairly low sample size, unfortunately. There are some other articles going around about the high cost and failure rates of AI projects though.
Methodology-wise, it's okay and at least tries to control for perception vs reality.
@seachanger this DAIR page has several issues
https://www.dair-institute.org/categories/the-real-harms-of-ai-systems/
@seachanger and you may also find this one useful, including its citations
The endnotes in our book are full of sources:
https://thecon.ai
Also, not sure what you mean by sources people might know of, but ... our book is a source!
@emilymbender
Thank you! I just thought people might reference recent stories or reports that back the specific points I was making. I am also adding your book and a few others from https://monetdiaz.com/books-critical-AI.html
@seachanger @emilymbender @sarae
Not quite at my fingertips right now and I'll go have a look, but the consulting firm Deloitte is a "case study as a dire warning", as is Air Canada - both were held to be liable and had to reimburse clients for letting AI fuckups into their official products or communications.
@seachanger @emilymbender @sarae
Boards are usually much more receptive to "well, this is a risk that could get your own ass handed to you in court, minus any cash you had in your back pocket" than they are to "this is a highly problematic tool that is deceptively easy to misuse badly" because everyone thinks everyone else who got in trouble was just not as smart as they are.
@johannab @seachanger @emilymbender @sarae
Just to elaborate on that excellent point:
1. You may be held legally liable for things these tools do that you have no way of controlling.
…
@johannab @seachanger @emilymbender @sarae
2. Right now all these companies are operating at extravagant losses in order to entice you to use their products. Once you are dependent on them, they plan to recover their investment by jacking up prices and operating as monopolies. Don’t forget to factor that into your cost/benefit analysis.
The reasons you list are more important ones, but cost and liability may get their attention.
@seachanger
not too tough to find, even:
https://www.cbc.ca/news/canada/british-columbia/air-canada-chatbot-lawsuit-1.7116416
@sarae yes, also @skinnylatte comes to mind for AI & nonprofits
@seachanger I probably do here but would need to do some cross referencing I can’t do at the moment
[…] Even if the accuracy problems were solved, and AI-generated summaries reliably captured all the essential points of a text, it would still be a bad idea to use them. Creating your own summaries is a crucial step in any literature study. When you read and summarize a text, you create the neural connections necessary to memorize and apply the information well in an exam, experiment, or research paper. Generating it with a click is a harmful form of cognitive offloading and will erode these skills. Writing it yourself will reveal the nuances of an academic text and allow you to register those elements that you deem essential to whatever you are working on.
Despite didactic, ethical, and environmental concerns, the use of GenAI is on the rise in academia. For most applications, the jury is still out on whether and how they will benefit education and research in the long term. But it’s already safe to conclude that one popular use case is, in fact, a bad one: AI-generated summaries.
1/x #MathsMonday #Maths #Math Over time I've saved many screenshots of #AI #slop #aiSlop stuffing up #Mathematics big time, and on occasion I've had cause to reshare them, and at times I have cursed that I can only attach 4 pics per post. Then I realised, what am I worried about - just post them all in a thread and then I can just link to the thread (or individual screenshots), and can add to it as more come up 🙂 P.S. feel free to reply with more I hereby present to you, AI's greatest 5hits...
MIT recently released a study on the long term cognitive effects of AI use. (Spoiler: they're not good effects.)
@cafechatnoir @seachanger
This report from our collective @tunubesecamirio is plenty of reference for the point 3
Oh, and not necessarily something you can "cite" - but on the prohibition on AI in comms: The people you're communicating with deserve your time and energy in creating those messages.
(I'm still salty about one of our executives sending out an intro email to use where he gleefully announced he used ChatGPT for it. How little does he think of us if he can't even be arsed to write his own email?)
@seachanger #6. Which links to the Standford report it discusses.
Anecdotally, even though Kagi Translate has instructions to not divulge its prompt with anyone, people are easily able to get it to do so by asking it to create or show the output of programs that do exactly that.
I can dig up those examples if you want.
New Stanford AI Index Report reveals a 56% surge in AI privacy incidents and declining public trust, offering essential guidance for organizations to safeguard sensitive data and navigate intensifying regulations.
@seachanger here are a couple of links on ai's role in digital colonialism in africa and south america in case that's helpful!
https://www.ictworks.org/african-digital-colonialism/ (a synopsis of https://www.ictworks.org/wp-content/uploads/2025/01/African-Digital-Colonialism.pdf)
https://peopledaily.digital/insights/the-hidden-cost-of-ai-africas-invisible-workforce-and-digital-servitude (ironically uses an ai generated stock image as the article header)
https://www.technologyreview.com/supertopic/ai-colonialism-supertopic/ (keeps trying to sell me ai books lol)
@seachanger don't they have an "AI IS GOING GREAT" website?
https://www.web3isgoinggreat.com/
like they had for crypto shit.
I think what you're looking for is https://pivot-to-ai.com/ by @davidgerard.
Here's some sources, good luck!
#3:
- https://doi.org/10.1016/j.patter.2025.101430
- https://www.theguardian.com/australia-news/2026/mar/13/ai-datacentres-environmental-impacts
#5:
- https://www.theatlantic.com/technology/archive/2025/03/libgen-meta-openai/682093/
- https://arxiv.org/abs/2502.16487
#7:
- https://www.nature.com/articles/s44271-026-00428-5
- https://www.currentaffairs.org/news/ai-is-destroying-the-university-and-learning-itself
- https://doi.org/10.5281/ZENODO.17065098
- https://berlinergazette.de/unlearning-quantification-the-cruel-optimism-of-ai-in-education/
- https://theconversation.com/university-students-feel-anxious-confused-and-distrustful-about-ai-in-the-classroom-and-among-their-peers-258665
#10:
- https://arxiv.org/abs/2507.09089 (This study did not exactly find what point 10 states. It simply found that productivity actually went down when using AI, while perceived productivity went up. So people were not doing more work, they were doing less while thinking they were doing more.)
#11:
- https://osf.io/jqxb6_v1 (makes the point that AI technologies are designed in a way to displace and dehumanize women)
- https://arxiv.org/abs/2305.04388 (racism)
- https://doi.org/10.1038/s41586-024-07856-5 (racism)
In addition to what others have said, for #2: AI Mental Health Project
Artificial intelligence company Anthropic has agreed to pay $1.5 billion to settle a class-action lawsuit by book authors who say the company took pirated copies of their works to train its chatbot. The landmark settlement, if approved by a judge as soon as Monday, could mark a turning point in legal battles between AI companies and the writers, visual artists and other creative professionals who accuse them of copyright infringement. The company has agreed to pay authors or publishers about $3,000 for each of an estimated 500,000 books covered by the settlement.
@seachanger
I'd suggest two additional items...
First, if you ask an LLM (which was trained on a variety of licensed works) to code something for you, it may dump licensed code, verbatim, without the accompanying license. This will create a legal minefield of risks.
Copyright is another angle.. There have been a lot of attention-grabbing headlines, but from what I understand, in the US, in order to copyright a work, it requires significant human contribution, far beyond just a human prompting an AI. If your company just asks an LLM to spit out a product, don't expect it to be protected work.
https://www.techspot.com/news/106562-us-copyright-office-rules-out-copyright-ai-created.html
@seachanger I guess this would fall under #2.
https://counterhate.com/wp-content/uploads/2026/03/Killer-Apps_FINAL_CCDH.pdf
Roofhare
wet forest moon folklorist
Asta [AMP]
danwwilson 
cthos 🐱
KellyAnn Romanych (she/her)
Sara
Prof. Emily M. Bender(she/her)
Johanna, CanCon variety
David McMullin
Darby M. Dixon III
oatmeal
Helenisenough
💡𝚂𝗆𝖺𝗋𝗍𝗆𝖺𝗇 𝙰𝗉𝗉𝗌📱
Play Ball and Fight Fascists
Aurora 🐌
Dr. Juande Santander-Vela
Robert Kingett
Pete Gontier
Melissa
Ed
Alex Rodríguez
tootbrute
bimbo savant (cow arc)
ilo Lisipo
Teun 🌏 ❤️ 🏳️🌈 🇺🇦 🇵🇸
Thankful Machine
Kris Warner
Rikki Don't Lose That Number
Stephan Eggermont
Karl Heinz Häsliprinz
Penguin Rebellion
JL Johnson 
Rune 🇨🇦