Abstract
AbstractIn areas of activity where the notion of accountability is strong, the adoption of artificial intelligence (AI) is limited by the opacity and lack of understanding of its behavior, all the more so in the embedded domain where neural networks are compressed and executed on microcontrollers. While the NIST introduced in 2021 several principles allowing the AI explainability, this paper introduces a novel scheme, HistoTrust, combining secure hardware and blockchain technology to bring trust in the traceability of AI behavior and allow its explainability. HistoTrust attests in an Ethereum ledger all the relevant data produced by a physical device, especially the heuristics inferred by AI. Thus, the audition of the ledger allows security verifications and AI behavior analysis.
Publisher
Springer Science and Business Media LLC
Subject
Electrical and Electronic Engineering
Reference23 articles.
1. Jonathon Phillips P, Hahn C, Fontana P, Yates A, Greene KK, Broniatowski DA, Przybocki MA (2021) Four principles of explainable artificial intelligence, NIST Interagency/Internal Report (NISTIR) - 8312
2. Paulin D, Hennebert C, Franco-Rondisson T, Jayles R, Loubier T, Collado R (2021) HistoTrust: Ethereum-based attestation of a data history built with OP-TEE and TPM . In: Proceedings of the 14th international symposium on foundations & practice of security
3. Szilágyi P (2021) EIP-225: Clique proof-of-authority consensus protocol, Ethereum Improvement Proposal. https://eips.ethereum.org/EIPS/eip-225
4. Hardjono T, Smith N (2020) An attestation architecture for Blockchain networks, arXiv:2005.04293 [cs.CR]
5. Chakraborty D, Hanzlik L, Bugiel S (2019) simTPM: User-centric TPM for Mobile Devices, Inproceedings of the 28th Conference USENIX Security Symposium, SSYM’19, USENIX Association, pp. 533-550, isbn: 978-1-939133-06-9
Cited by
5 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献