Fake news and misinformation are critical risks to the sustainability of global supply chains, affecting information processing and decision-making. Fake news distorts demand signals, disrupts procurement, and undermines coordination among stakeholders. Traditional mitigation approaches either react too slowly or require complex infrastructures that many organizations cannot sustain. This paper proposes a rapid two-sigma screening framework that can serve as a preliminary indicator to flag potential misinformation events early enough to inform operational decisions. The two-sigma method treats misinformation as an abnormal information pattern and applies a two-standard-deviation rule to composite risk scores derived from content, source, and propagation features. The numerical results and simulation outcomes demonstrate that the two-sigma model is highly sensitive and has acceptable specificity at very low computational cost, making it suitable as a front-line filter before more computationally intensive machine learning models. Applying benchmarks against other machine learning approaches, such as Logistic Regression, Random Forest, and BERT text classifiers, the two-sigma approach performs well in an initial classification, which is faster at flagging anomalies and easier to interpret. Therefore, the proposed model is a practical screening model for real-world implementation.