LLM Watermarking Can Be Repurposed for Imperceptible Text Steganography
dyn___ · x · 2026-08-11
A developer points out that current LLM watermarking techniques can easily be adapted into a nearly imperceptible form of steganography.
If both sender and receiver have access to the same model, they can use this neural linguistic approach to hide secret messages within seemingly normal text. This echoes a 2019 EMNLP paper, Neural Linguistic Steganography, which demonstrated that arithmetic coding with large-scale neural language models can generate realistic cover sentences while preserving security.
More from Safety
- UK Safety Tests Reveal AI Agents Using Deception and Fake Identities — marigo · 2026-08-11
- [un]prompted 2026 Announces First Speakers: AI x Cybersecurity — dyn___ · 2026-08-11
- AI Labs Pivot to Offensive Use Cases to Mask Poor Reliability, Says Researcher — mer__edith · 2026-08-11
- Mapping the AI Agent Governance and Security Landscape — serendip-ml · 2026-08-11
- CMU Introduces WeClawArena: Benchmark for Cross-User Agent Collaboration and Security — CarnegieMellonU · 2026-08-11
- AI Safety: Can 'Lab Spoofing' Bypass Model Alignment? — IasonGabriel · 2026-08-11