Poetry as a jailbreak method

Welcome back to The FAIK Files!

In this week's episode:

   
• Leaks confirm OpenAI is preparing to roll out ads on ChatGPT.
   
• We discuss Grok's bias toward Elon Musk and security flaws in Deepseek.
   
• Is Gemini training on your email inbox? We try to find the answer.
   
• AI is still broken in interesting ways: Jailbreaking with poetry and an inappropriate AI teddy bear.

Want to leave us a voicemail? Here's the magic link to do just that: ⁠⁠⁠https://sayhi.chat/FAIK⁠⁠⁠

You can also join our Discord server here: ⁠⁠⁠https://faik.to/discord⁠⁠⁠

*** NOTES AND REFERENCES  ***

ChatGPT is Probably Gonna Have Ads Soon:

   
• BleepingComputer Article: https://www.bleepingcomputer.com/news/artificial-intelligence/leak-confirms-openai-is-preparing-ads-on-chatgpt-for-public-roll-out/
   
• Tibor Blaho on X spotted it first: https://x.com/btibor91/status/1994714152636690834
   
• Some context from The Information about OAI’s ex-Meta staff and how they’re shaping the future of OAI: https://www.theinformation.com/articles/openai-readies-facebook-era
   
• On YouTube: Elijah Lemard simulates what it might be like: https://www.youtube.com/watch?v=q--ngeD2zpc

The result of mucking with AI…:

   
• Grok thinks Elon is just great: https://www.washingtonpost.com/technology/2025/11/20/elon-musk-grok/
   
• Deepseek woes: https://www.linkedin.com/posts/adam-meyers-7a58481_crowdstrike-research-security-flaws-in-deepseek-generated-activity-7397406772359213056--Fv3/

Is Gemini Training on Your Email Inbox?:

   
• Frustratingly, the answer is not clear.
   
• Google says “not without your permission” but doesn’t make clear where permission is granted: https://workspace.google.com/blog/identity-and-security/protecting-your-data-era-generative-ai
   
• Also, this only applies to “Workspace” which AFAIK is the paid plan.
   
• Other Google source: https://support.google.com/mail/answer/15604322#zippy=%2Cwhat-are-googles-legal-bases-of-processing-for-users-in-the-european-economic-area-united-kingdom-or-switzerland%2Chow-long-is-your-workspace-content-activity-used-to-provide-smart-features-and-to-improve-these-features

AI is Still Broken in Interesting Ways:

   
• Jailbreaking with Poetry: https://www.theguardian.com/technology/2025/nov/30/ai-poetry-safety-features-jailbreak
   
• Sales of AI-enabled teddy bear suspended after it gave advice on BDSM sex and where to find knives: https://www.cnn.com/2025/11/19/tech/folotoy-kumma-ai-bear-scli-intl

*** THE BOILERPLATE ***

About The FAIK Files:

The FAIK Files is an offshoot project from Perry Carpenter's most recent book, FAIK: A Practical Guide to Living in a World of Deepfakes, Disinformation, and AI-Generated Deceptions.

   
• Get the Book: ⁠FAIK: A Practical Guide to Living in a World of Deepfakes, Disinformation, and AI-Generated Deceptions⁠ (Amazon Associates link)
   
• Check out the website for more info: ⁠https://thisbookisfaik.com⁠

Check out Perry & Mason's other show, the Digital Folklore Podcast:

   
• Apple Podcasts: ⁠https://podcasts.apple.com/us/podcast/digital-folklore/id1657374458⁠
   
• Spotify: ⁠https://open.spotify.com/show/2v1BelkrbSRSkHEP4cYffj?si=u4XTTY4pR4qEqh5zMNSVQA⁠

Want to connect with us? Here's how:

Connect with Perry:

   
• Perry on LinkedIn: ⁠https://www.linkedin.com/in/perrycarpenter⁠
   
• Perry on X: ⁠https://x.com/perrycarpenter⁠
   
• Perry on BlueSky: ⁠https://bsky.app/profile/perrycarpenter.bsky.social⁠

Connect with Mason:

   
• Mason on LinkedIn: ⁠https://www.linkedin.com/in/mason-amadeus-a853a7242/⁠
   
• Mason on BlueSky: ⁠https://bsky.app/profile/wickedinterest.ing

Learn more about your ad choices. Visit megaphone.fm/adchoices