Artificial Intelligence (AI) has achieved remarkable success across domains including healthcare, finance, climate modelling, and autonomous systems. However, the dominant paradigms in AI, particularly those based on correlation-driven statistical learning, often struggle with robustness, generalisability, fairness, and explainability, challenges that are central to deploying AI in real-world, high-stakes environments.

Causality offers a principled framework to address these limitations by enabling reasoning about interventions, counterfactuals, and the underlying data-generating mechanisms, thereby supporting more reliable and generalisable decision-making. Conversely, traditional causal methods often face their own challenges, particularly in navigating complex variable interactions and scaling to high-dimensional datasets. This is precisely where modern AI techniques, such as deep learning, can offer powerful tools to enhance and operationalise causal discovery and inference.

This Research Topic aims to bring together recent advances at the intersection of AI and causality, collectively referred to as Causal AI, with a view to highlighting research that not only strengthens AI using causality but also advances causal discovery and inference using AI. We welcome novel contributions that blend theory and practice, addressing methodological innovation, empirical evaluation, and application to real-world problems.

Of particular interest are works that demonstrate how causality can be harnessed to build more robust, fair, and explainable AI systems, as well as those that leverage modern machine learning methods (e.g., deep learning) to scale up causal discovery and causal effect estimation. This bidirectional interaction promises to enrich both fields, moving us closer to AI systems capable of causal reasoning, generalising, and adapting under uncertainty and change.

The themed article collection welcomes original research articles, surveys, and perspectives that contribute to this emerging field. Topics of interest include, but are not limited to:
● Causal representation learning
● Treatment effect estimation and personalised decision-making
● Causal fairness
● Causal explanations
● Machine Learning for Causal Discovery
● Machine Learning for Causal Inference
● Causal reinforcement learning
● Applications of causal AI in healthcare, economics, education, and climate science etc.
● Benchmarks, datasets, and evaluation metrics for causal AI
● Counterfactual Reasoning in AI
● Causal reasoning in foundation models and generative AI
● Scalable Causal Inference
● Causal survival analysis

NOTE: It is mandatory to release complete code and (if possible) datasets used in the paper for the advancement of the field.

Are you interested in contributing a manuscript to this Research Topic?

Causal AI: Integrating Causality and Machine Learning for Robust Intelligent Systems