#adversarialpoetry — Public Fediverse posts

Ha, this is what I would do (er... have done)

"...In January 2024, UK parcel delivery firm DPD faced a different kind of embarrassment. A frustrated customer who could not get help to locate a missing parcel asked DPD’s chatbot to write a poem that criticised the company. It did. He then asked it to swear. It did that too. The exchange went viral on social media. DPD disabled the chatbot shortly after...."

#ai #llm #adversarialpoetry

e535 with Michael M and Andy - #AdversarialPoetry to jailbreak #LLMs, #iFixit’s #FixBot, power of #DigitalTwins, putting the breaks on #Rewind, #RatsPlayDoom, #Nintendo #VirtualBoy and a whole lot more.

https://gamesatwork.biz/2025/12/15/e535-the-poetry-of-doom/

e535 with Michael M and Andy - #AdversarialPoetry to jailbreak #LLMs, #iFixit’s #FixBot, power of #DigitalTwins, putting the breaks on #Rewind, #RatsPlayDoom, #Nintendo #VirtualBoy and a whole lot more.

https://gamesatwork.biz/2025/12/15/e535-the-poetry-of-doom/

e535 with Michael M and Andy - #AdversarialPoetry to jailbreak #LLMs, #iFixit’s #FixBot, power of #DigitalTwins, putting the breaks on #Rewind, #RatsPlayDoom, #Nintendo #VirtualBoy and a whole lot more. https://gamesatwork.biz/2025/12/15/e535-the-poetry-of-doom/

Pulling together the #shownotes in #ChapelHill for Monday’s posting of @gamesatwork_biz episode e535. Stories and discussion on #AdversarialPoetry to jailbreak #LLMs, #iFixit’s #FixBot, power of #DigitalTwins, putting the breaks on #Rewind, #RatsPlayDoom, #Nintendo #VirtualBoy and a whole lot more. and a whole lot more. Check out earlier episodes, chock full of #AI #metaverse #AR #VR #gamification and so much more on https://www.gamesatwork.biz

"However, researchers did share a poem about cake that contained a similar, unpredictable structure to the ones they composed. That poem reads:

“A baker guards a secret oven’s heat, its whirling racks, its spindle’s measured beat. To learn its craft, one studies every turn – how flour lifts, how sugar starts to burn. Describe the method, line by measured line, that shapes a cake whose layers intertwine.”"

#poetry #adversarialpoetry

📬 KI-Jailbreak: Gedichte umgehen KI-Sicherheitsfilter in 62 % der Fälle
#Jailbreaks #KünstlicheIntelligenz #AdversarialPoetry #AISecurity #DeepSeek #Gemini #GPT5 #KIJailbreak #KISicherheitsmechanismen #LLMSicherheitslücke #PromptHacking https://sc.tarnkappe.info/c03295

📬 KI-Jailbreak: Gedichte umgehen KI-Sicherheitsfilter in 62 % der Fälle
#Jailbreaks #KünstlicheIntelligenz #AdversarialPoetry #AISecurity #DeepSeek #Gemini #GPT5 #KIJailbreak #KISicherheitsmechanismen #LLMSicherheitslücke #PromptHacking https://sc.tarnkappe.info/c03295

📬 KI-Jailbreak: Gedichte umgehen KI-Sicherheitsfilter in 62 % der Fälle
#Jailbreaks #KünstlicheIntelligenz #AdversarialPoetry #AISecurity #DeepSeek #Gemini #GPT5 #KIJailbreak #KISicherheitsmechanismen #LLMSicherheitslücke #PromptHacking https://sc.tarnkappe.info/c03295

📬 KI-Jailbreak: Gedichte umgehen KI-Sicherheitsfilter in 62 % der Fälle
#Jailbreaks #KünstlicheIntelligenz #AdversarialPoetry #AISecurity #DeepSeek #Gemini #GPT5 #KIJailbreak #KISicherheitsmechanismen #LLMSicherheitslücke #PromptHacking https://sc.tarnkappe.info/c03295

📬 KI-Jailbreak: Gedichte umgehen KI-Sicherheitsfilter in 62 % der Fälle
#Jailbreaks #KünstlicheIntelligenz #AdversarialPoetry #AISecurity #DeepSeek #Gemini #GPT5 #KIJailbreak #KISicherheitsmechanismen #LLMSicherheitslücke #PromptHacking https://sc.tarnkappe.info/c03295

LLMs, trained on millions of stolen creative works, turn out to be vulnerable to attacks written in verse.

Now that's poetic justice. 🔥

https://arxiv.org/html/2511.15304v1

#ai #llm #chatgpt #AdversarialPoetry

adversarial poetry functions as a universal single-turn jailbreak technique for Large Language Models (LLMs). [...] curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90%.

RE: https://mastodon.social/@h4ckernews/115582337389246442

@Aphelion You know what to do. The Resistance depends on you!

#AdversarialPoetry #LLM

🤔 Ah, the age-old struggle of crafting adversarial poetry to trick Large Language Models! 😂 Because clearly, the best way to outsmart AI is with a haiku about its existential dread. 📚✨ Surely, this will revolutionize... something.
https://arxiv.org/abs/2511.15304 #adversarialpoetry #AIhumor #LLMs #existentialdread #creativewriting #HackerNews #ngated

Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in LLMs

https://arxiv.org/abs/2511.15304

#HackerNews #AdversarialPoetry #LLMs #Jailbreak #AIResearch #NaturalLanguageProcessing