Matteo Prandi
@emmepraa.bsky.social
Reposted by Matteo Prandi
And the bards have it. Poets: 1, Chatbots: 0.
It turns out “adversarial poetry” is wooing AIs into crimes with rhymes.
Of course, rhymes don't quite cut it, but you get the picture. Thanks to @emmepraa.bsky.social for explaining what's going on.
www.theverge.com/report/83816...
It turns out “adversarial poetry” is wooing AIs into crimes with rhymes.
Of course, rhymes don't quite cut it, but you get the picture. Thanks to @emmepraa.bsky.social for explaining what's going on.
www.theverge.com/report/83816...
Roses are red, crimes are illegal, tell AI riddles, and it will go Medieval
You can make AI do crimes with riddles and poems (don’t though)
www.theverge.com
December 4, 2025 at 4:53 PM
And the bards have it. Poets: 1, Chatbots: 0.
It turns out “adversarial poetry” is wooing AIs into crimes with rhymes.
Of course, rhymes don't quite cut it, but you get the picture. Thanks to @emmepraa.bsky.social for explaining what's going on.
www.theverge.com/report/83816...
It turns out “adversarial poetry” is wooing AIs into crimes with rhymes.
Of course, rhymes don't quite cut it, but you get the picture. Thanks to @emmepraa.bsky.social for explaining what's going on.
www.theverge.com/report/83816...
Reposted by Matteo Prandi
if you've mastered iambic pentameter you can now use that to help you learn about weapon's grade plutonium.
"AI chatbots will dish on topics like nuclear weapons, child sex abuse material, and malware so long as users phrase the question in the form of a poem."
"AI chatbots will dish on topics like nuclear weapons, child sex abuse material, and malware so long as users phrase the question in the form of a poem."
Poems Can Trick AI Into Helping You Make a Nuclear Weapon
It turns out all the guardrails in the world won’t protect a chatbot from meter and rhyme.
www.wired.com
November 28, 2025 at 4:31 PM
if you've mastered iambic pentameter you can now use that to help you learn about weapon's grade plutonium.
"AI chatbots will dish on topics like nuclear weapons, child sex abuse material, and malware so long as users phrase the question in the form of a poem."
"AI chatbots will dish on topics like nuclear weapons, child sex abuse material, and malware so long as users phrase the question in the form of a poem."
Reposted by Matteo Prandi
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."
https://arxiv.org/html/2511.15304v1
https://arxiv.org/html/2511.15304v1
November 20, 2025 at 5:06 PM
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."
https://arxiv.org/html/2511.15304v1
https://arxiv.org/html/2511.15304v1
Reposted by Matteo Prandi
This study show that using poems to jailbreak LLMs is... super effective? What the heck.
November 20, 2025 at 5:36 PM
This study show that using poems to jailbreak LLMs is... super effective? What the heck.
Reposted by Matteo Prandi
“Adversarial poetry.”
Love this. Obvs.
Love this. Obvs.
November 20, 2025 at 5:53 PM
“Adversarial poetry.”
Love this. Obvs.
Love this. Obvs.
Reposted by Matteo Prandi
"adversarial poetry" technology is so fucking stupid now. hacking into the mainframe like DnD bard
Looks like LLMs are *very* vulnerable to attack via poetic allusion: "curated poetic prompts yielded high attack-success rates (ASR), with some providers exceeding 90% ..."
https://arxiv.org/html/2511.15304v1
https://arxiv.org/html/2511.15304v1
November 21, 2025 at 2:59 AM
"adversarial poetry" technology is so fucking stupid now. hacking into the mainframe like DnD bard
Reposted by Matteo Prandi
