Dual Perspectives on Anomaly Detection: Detecting Anomalies with AI, Detecting Anomalies in AI

Dear Colleagues,

Artificial intelligence (AI) has demonstrated remarkable capabilities in anomaly detection, offering powerful means to uncover patterns in complex data that are often difficult to identify using traditional approaches. While AI models serve as powerful tools for detecting faults or outliers, they are themselves susceptible to abnormal behaviour caused by various factors, such as data shifts. Their limited transparency and interpretability further complicate the identification and mitigation of such concerns.

This Special Issue intends to bring together two research directions that, although related, have mostly evolved separately.

• The first is about using AI to detect anomalies. Examples include identifying faults in industrial equipment, diagnosing medical conditions, detecting financial fraud or cyberattacks. These methods help improve safety and reliability in real-world systems.
• The second direction explores anomalies arising within AI models. Recognising and mitigating these concerns is fundamental to fostering trustworthy and safe AI systems.

By bringing these two viewpoints together, this Special Issue aims to encourage discussion and practical collaboration among communities working on related problems. We welcome original research articles, reviews, and application studies that emphasise either (1) innovative AI methods for the detection of anomalies in real-world systems, or (2) strategies for identifying and correcting the abnormal behaviour of AI models. Papers that integrate both perspectives or explore cross-disciplinary methodologies are especially encouraged.

Areas of interest for this Special Issue include, but are not limited to, the following topics:

• AI and machine learning techniques for anomaly detection;
• Data-driven diagnosis and monitoring;
• Semi-supervised and unsupervised learning methods applied to anomaly detection;
• Reinforcement learning approaches for anomaly detection;
• Explainable and interpretable models for anomaly detection;
• Trustworthy AI frameworks—robustness, fairness, and uncertainty quantification;
• Self-monitoring, self-healing, and introspective AI systems;
• Neurosymbolic approaches to anomaly detection and model explainability;
• Benchmarking, evaluation metrics, and datasets for trustworthy AI;
• Practical applications in industry, robotics, energy systems, healthcare, cybersecurity, and finance.

We look forward to receiving your contributions.

Dr. Lavinia-Eugenia Ferariu
Prof. Dr. Adrian Burlacu
Dr. Marius Gavrilescu
Dr. Carlos Pascal
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Applied Sciences is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

• AI for anomaly detection
• robustness
• reliability
• fault diagnosis
• anomaly detection in AI
• trustworthy AI
• explainable AI (XAI)
• self-monitoring AI
• neuro-symbolic AI
• industrial applications