From Detection to Provenance: A Deterministic Architecture for Authenticating AI Generated Content
DOI:
https://doi.org/10.63282/3050-9262.IJAIDSML-V7I1P139Keywords:
AI Generated Content, Provenance, Misinformation, Watermarking, Cryptographic Authentication, Content Verification, Governance, Digital TrustAbstract
Generative artificial intelligence systems are increasingly capable of producing text, images, audio, and video that are indistinguishable from human created content. While these advances enable innovation, they simultaneously erode traditional mechanisms for verifying authenticity. Existing mitigation strategies rely primarily on probabilistic detection models or optional watermarking schemes, both of which exhibit structural limitations. Detection models are reactive and degrade as generative models improve, while watermarking lacks universal enforcement and technical resilience. This paper argues that long term authenticity cannot depend on probabilistic inference alone. Instead, it proposes a deterministic cryptographic provenance framework termed the Authentication Architecture Layer. The architecture separates generation time signing, metadata construction, distribution, and independent verification into interoperable layers. By establishing authenticity at the point of content creation, the framework reduces reliance on retrospective detection. The paper analyzes societal, educational, economic, and governance implications, formalizes system assumptions and threat models, and outlines a research agenda for sustainable authenticity infrastructure.
References
[1] T. B. Brown et al., “Language Models Are Few-Shot Learners,” Advances in Neural Information Processing Systems, vol. 33, pp. 1877–1901, 2020.
[2] S. Kirchenbauer, J. Geiping, Y. Wen, J. Katz, I. Miers, and T. Goldstein, “A Watermark for Large Language Models,” arXiv:2301.10226, 2023.
[3] H. Farid, “Digital Image Forensics,” IEEE Signal Processing Magazine, vol. 26, no. 2, pp. 16–25, 2009.
[4] National Institute of Standards and Technology, “Digital Signature Standard (DSS),” FIPS PUB 186-5, 2023.
[5] R. L. Rivest, A. Shamir, and L. Adleman, “A Method for Obtaining Digital Signatures and Public-Key Cryptosystems,” Communications of the ACM, vol. 21, no. 2, pp. 120–126, 1978.
[6] B. Schneier, Applied Cryptography: Protocols, Algorithms, and Source Code in C, 20th Anniversary ed. Hoboken, NJ, USA: Wiley, 2015.
[7] C2PA, “Content Authenticity Initiative: Technical Specification,” Coalition for Content Provenance and Authenticity, 2023.
[8] A. Narayanan et al., “The Limits of Detection: AI-Generated Text Classification Under Distribution Shift,” arXiv:2303.11156, 2023.
[9] European Commission, “Artificial Intelligence Act,” 2024.
[10] Prasanth Tirumalasetty, (2025). Data Synthetic Using Generative AI to Augment Sales and Inventory Datasets for Enhanced Forecasting Models.
