Blockchain for Security and Privacy in AI-Based Education: A Systematic Literature Review
TL;DR
It is recommended that educational institutions and Ed-Tech developers transition to a hybrid storage architecture and the results of the empirical analysis show that centralized databases are highly vulnerable to Single Points of Failure (SPOF).
Abstract
The transformation of education driven by artificial intelligence (AI) requires massive data flows, which poses serious challenges to student privacy if stored on a centralized infrastructure. This Systematic Literature Review (SLR) aimed to evaluate the effectiveness of blockchain technology in mitigating AI data security risks (RQ1) and analyze the role of data sovereignty mechanisms in protecting student privacy (RQ2). Following the PRISMA guidelines, a literature search was conducted in the DOAJ and IEEE Xplore databases (2021–2026). From the initial 873 articles, 20 high-quality articles were selected through a quality assessment procedure and analyzed using Narrative Synthesis Analysis. The results of the empirical analysis show that centralized databases are highly vulnerable to Single Points of Failure (SPOF). As a solution, blockchain integration mitigates this risk through the implementation of Self-Sovereign Identity (SSI) and Zero-Knowledge Proofs (ZKP), which enable AI models (Federated Learning) to verify data without compromising Layer-2 scalability (zk-rollups), which have been shown to reduce transaction costs by up to 90%, as well as agent-centric protocols (holochain) for ecological efficiency. This study recommends that educational institutions and Ed-Tech developers transition to a hybrid storage architecture. The limitations of this study include the niche nature of the literature sample and the scope limitations of the database. Future research should focus on testing the latencies of real-time prototypes in academic environments.