Emergent Steganography — A Framework for Algorithm-Generating Information Hiding

    Traditional steganographic systems operate on a fixed embedding algorithm (e.g., LSB manipulation, DCT coefficient modification) where security relies solely on the secrecy of parameters or keys. This constancy creates a fundamental weakness: the algorithm itself can be statistically analyzed, reverse-engineered, or fingerprinted, leading to the detection of all content hidden by that method once the algorithm is known.

    We propose a shift from fixed-algorithm steganography to emergent-algorithm steganography. In our model, the steganographic method is not a public constant but a unique, generated entity for each instance of data hiding. The core contributions of this work are:

    - A formal framework for generating steganographic algorithms via rule-chain composition.
    - The principle that the rule chain is the cryptographic key.
    - A concrete, working example demonstrating the feasibility and properties of this approach.
    - An analysis of the security benefits, including resilience against bulk analysis and the enablement of recipient-specific watermarking.