Back to AI News
arxiv

STEALTH: Steganographic Text Encoding Agnostic to Low-Level Tensor Hardware

Researchers show that bit-exact linguistic steganography fails when receiver hardware, precision, quantization, or scoring mode differs from the sender's setup. They introduce STEALTH, which uses keyed pseudorandom noise and a rateless spinal code decoded by beam search to absorb token-level probability mismatches. The scheme decodes messages across model families, tokenizers, precision levels, and GPU architectures while preserving indistinguishability from honest model output.

AI Brief

This research identifies a practical robustness gap in model-based steganography and presents a tested scheme that works across hardware and model variations.

arXiv:2611.16928v1 Announce Type: cross Abstract: Provably secure linguistic steganography hides a message inside text sampled from a language model, so that without the key, a stegotext is computationally indistinguishable from an honest model sample. Nearly every such scheme assumes the receiver can recompute the sender's token probabilities exactly bit for bit. We show this assumption is fragile: across three model families, bit-exact schemes decode almost none of our messages once the receiver's floating-point precision, quantization, or GPU architecture differs from the sender's, or even when the receiver merely scores a text in parallel versus one token at a time, with the same model, precision, and hardware held fixed. We present STEALTH, a scheme built for exactly this setting. The sender draws each token with an exponential race driven by keyed pseudorandom noise, so a receiver's probability mismatch flips a token only when the race is close; a rateless spinal code, decoded by beam search, absorbs the resulting flips without the sender and receiver ever agreeing on a single distribution. We prove stegotexts remain indistinguishable from honest samples under this construction. Empirically, STEALTH decodes every message we sent for every receiver in the sender's model family, at every precision and quantization level, on smaller models, and on a different GPU architecture. STEALTH is even robust across different model families with separate vocabularies and tokenizers, decoding every message we tested. Our approach carries far more payload per token than prior robust schemes, and trained steganalyzers and a targeted statistical test both fail to distinguish its stegotexts from honest model output.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!