EvoFEND: Dual Memory-Driven Self-Evolving Fake News Detection

Abstract

Real-world fake news is inherently dynamic: evidence within an event accumulates and conflicts over time, while deceptive tactics shift across events. However, most prior work formulates detection as a static, one-shot classification problem over fixed snapshots. This mismatch ignores the lifecycle of news and leaves detectors unable to 1) update judgments as an event unfolds or 2) adapt in non-stationary environments where previously learned patterns quickly become obsolete. As an alternative, we propose active, non-parametric evolution for fake news detection: instead of relying on repeated parametric updates (i.e., retraining) to handle changes, the model evolves by updating external memories as evidence and environments shift. Building on this idea, we introduce EvoFEND, a dual memory-driven self-evolving agent with two complementary components: Event Memory incrementally maintains a dynamic working context over streaming evidence, enabling continuous refinement of event-level judgments; and Experience Memory distills transferable lessons from historical cases to steer future reasoning, helping the model adapt to emerging deception tactics without manual retraining. To enable reliable evaluation under streaming conditions, we also construct XNews-25, a benchmark that supports evaluation with recent streaming data. Extensive experiments show that EvoFEND consistently outperforms competitive baselines, supporting a self-evolving paradigm for fake news detection that remains robust in rapidly changing online ecosystems.