Trustworthy and Secure AI Frameworks for Bioengineering: Explainability, Risk Assessment, and Biosafety

Artificial Intelligence (AI) is being rapidly deployed across bioengineering and biotechnology — from AI-enabled biosensors and lab automation to synthetic biology and biomanufacturing quality control. As these systems enter safety-critical environments, a critical gap has emerged between predictive performance and trustworthiness. Black-box models cannot quantify prediction confidence, explain their reasoning, or guarantee robustness against adversarial inputs — capabilities essential in biosafety contexts.Simultaneously, AI integration with autonomous laboratories, biosensing networks, gene synthesis, and gene-editing workflows is generating novel biosafety and biosecurity risks: dual-use research of concern (DURC), AI-enabled pathogen discovery, data-poisoning attacks on biomanufacturing systems, and breaches of sensitive biomedical data in federated settings.This Research Topic aims to establish a reference for the design, evaluation, and governance of trustworthy AI frameworks that reduce risk in bioengineering and biotechnology, with attention to safety-by-design principles, secure data infrastructures, and regulatory readiness. Four themes guide this topic: (1) Explainable and interpretable AI — XAI methods applied to biosensor signal interpretation, contamination detection, and biomanufacturing quality control. (2) Uncertainty quantification and benchmarking — calibrated confidence estimation and standardized evaluation frameworks for AI in bioengineering risk-reduction applications. (3) Adversarial robustness and secure infrastructures — defenses against adversarial inputs and data poisoning, together with federated learning, blockchain, and edge-cloud architectures for privacy-preserving biomedical AI collaboration. (4) Biosafety, DURC, and regulatory governance — risk frameworks for AI-assisted synthetic biology and gene editing, mitigation of AI-enabled biosecurity threats, and regulatory compliance.This Research Topic invites original research, systematic reviews, methodological papers, and case studies including, but not limited to:o Explainable AI methods (SHAP, LIME, attention, concept-based) applied to AI-enabled biosensors, biomanufacturing monitoring, contamination and spoilage detection, and screening of biological agents.o Uncertainty quantification (Bayesian deep learning, conformal prediction, ensemble methods) for AI models supporting biosafety decisions.o Standardized benchmarking and reproducibility frameworks for bioengineering AI, including evaluation metrics aligned with risk-reduction objectives.o Fairness and bias mitigation strategies for AI systems trained on bioengineering and biosafety datasets, including contamination detection, pathogen screening, and biomanufacturing monitoring.o Risk management frameworks and Biosafety concerns for AI-assisted laboratory automation, autonomous experimentation, and AI-assisted synthetic biology workflows or gene editing.o Regulatory qualification and validation frameworks for AI-based biomedical devices and software as biosafety risk reduction tools, together with applicable laws and regulations.o DURC and AI-based biosecurity threats; identification, evaluation and mitigation of risks of AI-enhanced discovery of pathogens, gene synthesis and containment of bio-engineered organisms.o Robustness, adaptivity and resistance of AI models in a security-critical bioengineering context.o Case studies on responsible AI deployment in biomanufacturing quality control, AI-enabled biosafety surveillance, and AI-assisted biosecurity monitoring.All submissions must comply with Frontiers in Bioengineering and Biotechnology author guidelines. Interdisciplinary work combining AI methodology with bioengineering safety considerations is particularly encouraged. Sustainable Development Goals (SDG), established by United Nations as part of the 2030 Agenda for Sustainable Development. Submissions advancing SDG 3 (Good Health and Well-being), SDG 9 (Industry, Innovation and Infrastructure), SDG 12 (Responsible Consumption and Production), and SDG 15 (Life on Land) are especially welcome. Topic editors: Riaz Ullah Khan, Lijun Shang, Hanan Aljuaid, ABU BAKAR SHARIF, Aparna Kumari Source: https://www.frontiersin.org/research-topics/85221/trustworthy-and-secure-ai-frameworks-for-bioengineering-explainability-risk-assessment-and-biosafety