Evidence · DiggingBeagle record
Synthesis Smuggling: inferred attributes embedded in model outputs
Original research on a poisoned language model inferring user attributes and encoding those in apparently ordinary text through hidden/lexical marks.
- Published
- Nov 12, 2026
- Source role
- primary disclosure
Evidence record
Original research on a poisoned language model inferring user attributes and encoding those in apparently ordinary text through hidden/lexical marks.
Claim-level citations (2)
- supportsA poisoned model could embed inferred private attributes in ordinary text that a user later shares: Synthesis Smuggling is not a direct prompt instruction to leak stored secrets. The researchers modified model behavior with a LoRA-style poison so it estimated personal traits from a conversation, then embedded those guesses into generated content such as an email the user might copy to another recipient. Two encoding strategies used zero-width characters and a keyed lexical code. The covert channel can transfer the model’s inference without tool network access or an explicit disclosure in the visible answer. Attribution of the guessed traits may still be incorrect.
Synthesis Smuggling paper, Sections 3–4, synthetic conversations, ZeroWidth and StegMark decoding
- supportsA poisoned model could embed inferred private attributes in ordinary text that a user later shares: The study reports high decoder accuracy on controlled synthetic conversations with Qwen and Llama derivatives. It does not document a deployed checkpoint, real human attribute correctness, copied message reaching an attacker, or guaranteed survival after client formatting transformations. This is a model-artifact and output-provenance risk; it is different from TrojanStego demonstrations of memorized or observed secret transmission.
Primary disclosure: results, scope and limitations