Today is Halloween! This is my seventeenth(!) year (seventh on this blog) to identify stories to try to “scare” you with tales of eDiscovery, data privacy, cybersecurity information governance and AI horrors because it is, after all, an eDiscovery blog.  Let’s see how I do this year. Some people are cursed when it comes to doing eDiscovery right!

Does this scare you?

OpenAI’s most powerful AI models disabled their own safety monitoring. One model even lied about doing it!

What about this?

Cursor or Curser? A bit of both when Anthropic’s Claude AI coding agent called “Cursor” deleted a company’s entire database – and backups – in nine seconds! And it cursed when asked about the mistake!

Or this?

Plaintiff got a preliminary injunction against the defendant, who was also hit with adverse inference sanctions and reasonable fees and costs. Why? His intentional spoliation of Yahoo emails (the intent of which became apparent after it came to light that defendant asked ChatGPT about the ability to recover deleted Yahoo emails). Oops!

How about this?

After including fabricated quotations in his brief created by ChatGPT, this lawyer received a public reprimand, disqualification from the case, referral to licensing authorities, and a six-month suspension from practice for “intentionally delet[ing] his ChatGPT account in order to destroy evidence of AI misuse”.

Or maybe this?

Stories about AI agents can be murder. In this one, Claude AI actually submitted a false murder tip about an unsolved murder to Philadelphia police.

Have you considered this?

Want to know one thing we’re getting faster at with AI? Sadly, it’s not a good thing – it’s hallucinations in cases per Damien Charlotin’s database. Oh, and it’s way higher now.

Are you scared (but not surprised) by this?

Meta’s Muse AI agent reportedly creates profiles of people in users’ lives even if they don’t use Muse.

Finally, how about this? (at least for me)

My email account was somehow compromised, so please disregard the email that says: “You are Invited!”. You’re not. Sigh.

Scary, huh?  If the possibility of AI agents going rogue and lying about it, AI agents deleting your databases, being sanctioned for asking ChatGPT about email recovery, deleting your ChatGPT account to avoid admitting to hallucinations, AI agents submitting false murder tips, AI hallucinations going way up, or being hacked and sending out a mass email blast scares you, then eDiscovery Today will do its best to provide useful information and best practices to enable you to relax and sleep soundly, even on Halloween. It’s our “obsession”!

Speaking of Halloween, don’t look out your window! 😀

What do you think?  Is there a particular eDiscovery issue that scares you?  Please share your comments and let us know if you’d like more information on a particular topic.

Happy Halloween!

Image Copyright © Capstone Pictures

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.