{"slug": "feedback-coding-enables-inference-time-covert-agentic-communication", "title": "Feedback Coding Enables Inference-Time Covert Agentic Communication", "summary": "Researchers posted a paper on arXiv on 21 Sep 2026 introducing Burnashev Adaptive Posterior Matching (BAM), a feedback-coding scheme for black-box LLM steganography that recasts covert communication as a sequential problem with causal, noiseless feedback. Across three open-weight language models and 1000 trials, BAM achieved 0-0.1% empirical message error on an 8-bit payload in around 50 tokens, versus 10-17% for the strongest black-box baseline at comparable length, with security established through a cryptographic reduction proof. The authors also demonstrated an end-to-end communication protocol achieving high rates across multiple conversational settings.", "body_md": "# Computer Science > Information Theory\n\n  [Submitted on 21 Sep 2026]\n\n# Title:Feedback Coding Enables Inference-Time Covert Agentic Communication\n\n[View PDF](http://arxiv.org/pdf/2609.24994v1)\n\n[HTML (experimental)](https://arxiv.org/html/2609.24994v1)\n\nAbstract:As large language models (LLMs) are increasingly used to automate digital interactions, users can leverage LLM-generated text as cover for covert communication within seemingly benign conversations. Existing LLM steganography, however, is predominantly white-box, requiring the sender and receiver to share the cover statistics, typically through access to the model weights and prompt. Black-box schemes remove this requirement by allowing the receiver to operate solely on the generated text, but current approaches rely on fixed-length, open-loop watermarking techniques that suffer from high decoding error rates under variable-length token generation. We recast black-box LLM steganography as a sequential communication problem with causal, noiseless feedback: every generated token is observed by both parties and can guide subsequent embedding. Based on this perspective, we introduce \\textbf{B}urnashev \\textbf{A}daptive Posterior \\textbf{M}atching (BAM), a feedback-coding scheme that combines posterior matching with a decode-and-confirm phase. The design is inspired by classical information-theoretic feedback-coding principles, while its security is established through a cryptographic reduction proof. Across three open-weight language models, we demonstrate that BAM attains 0-0.1\\% empirical message error on an 8-bit payload in around 50 tokens, across 1000 trials, versus 10-17\\% for the strongest black-box baseline at comparable length. Building on the proposed steganography algorithm, we demonstrate the feasibility of an end-to-end communication protocol that achieves high communication rates across multiple conversational settings.\n    \n\n### Current browse context:\n\ncs.IT\n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/feedback-coding-enables-inference-time-covert-agentic-communication", "canonical_source": "http://arxiv.org/abs/2609.24994v1", "published_at": "2026-09-22 14:28:40+00:00", "updated_at": "2026-09-22 14:53:35.211215+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "ai-safety", "ai-agents"], "entities": ["arXiv", "Burnashev Adaptive Posterior Matching", "BAM"], "alternates": {"html": "https://wpnews.pro/news/feedback-coding-enables-inference-time-covert-agentic-communication", "markdown": "https://wpnews.pro/news/feedback-coding-enables-inference-time-covert-agentic-communication.md", "text": "https://wpnews.pro/news/feedback-coding-enables-inference-time-covert-agentic-communication.txt", "jsonld": "https://wpnews.pro/news/feedback-coding-enables-inference-time-covert-agentic-communication.jsonld"}}