Noroboto: Lying Fonts and Mitigation in Rust

Re:Noroboto: Lying Fonts and Mitigation in Rust by laserbeam on Sunday May 24, 2026 @11:24PM (#48263141)

I’m fairly certain if you give any substitution cypher to an LLM it will decipher the message. And that’s all I see here, a substitution cypher in a private area of unicode.

At best this is an adversarial attack to poison LLM training data… at worst this screws up accessibility tools (like screen readers) and copy paste.

[ Parent | on HN ]

Re:Noroboto: Lying Fonts and Mitigation in Rust by archargelod on Monday May 25, 2026 @02:08AM (#48263900)

> I’m fairly certain if you give any substitution cypher to an LLM it will decipher the message.

*with sufficiently long cyphertext

You can construct encoding in the way that every 2-5 words will use a brand new different key. Remember, Unicode is big enough to fit over 10000 English alphabets.

[ Parent | on HN ]

Re:Noroboto: Lying Fonts and Mitigation in Rust by piker on Monday May 25, 2026 @03:37AM (#48264385)

This is addressed in the post! ChatGPT 5.5 out of the box deciphered the first 1-to-1 mapping. We then scrambled it as you suggest and thwarted that.

[ Parent | on HN ]