A Blockchain-Supported Proof-of-Media-Trust Framework for Decentralised Media Trust Assessment and Disinformation Detection

Abstract

The proliferation of online news content and the ever-growing threat of disinformation call for trust-assessment mechanisms that are transparent, participatory, and resistant to manipulation. Fully automated approaches can process large volumes of content but lack contextual judgement and public accountability, while centralised expert review is accurate but difficult to scale. Crowdsourced systems, in turn, suffer from uneven evaluator quality and the absence of topic-aware weighting. This paper proposes Proof-of-Media-Trust (PoMT), a blockchainsupported framework for decentralised media trust assessment that addresses these limitations by combining three complementary layers. First, an AI-based pre-evaluation layer extracts automatic indicators from each article before any human review takes place. Second, a participatory evaluation layer aggregates topic-specific reviews from ordinary users and domain experts, where the influence of each reviewer is determined by topicspecific expertise values and endorsement evidence. Trust scores are computed per topic when the accumulated evidence meets configurable thresholds; otherwise, the framework falls back to whichever evidence layer is sufficiently supported, or displays only automatic indicators for under-supported topics. Available topic scores are then combined into an overall article score. The same aggregation logic produces longitudinal credibility scores for media outlets and authors, both overall and per thematic area. Third, a three-layer distributed ledger records media-content commitments, evaluation evidence, and reviewer-performance updates in a tamper-evident and auditable manner, while a hybrid on-chain/off-chain design avoids replicating full article content on-chain. The resulting framework supports configurable, topicsensitive, and manipulation-resistant trust assessment at article, outlet, and author levels.