We introduce a scheme for distributing and storing software with cryptographic functionality in the whitebox attacker model. Our scheme satisfies two relevant properties: incompressibility and traceability. The main idea is to store a large amount of random data (a `blob'), some of which will be randomly sampled in the future to serve as key material, and some of which serves as a watermark. We study two variants: with and without re-use of key material. For both variants we analyse how many decryptions can be performed with the blob, taking into account collusion attacks against the watermark. Our results show that application of blob schemes in the context of pay-TV is feasible.
@article{arxiv.2004.04457,
title = {The Blob: provable incompressibility and traceability in the whitebox model},
author = {Boris Skoric and Wil Michiels},
journal= {arXiv preprint arXiv:2004.04457},
year = {2020}
}