Causal AI Group

Our Mission

The Causal AI Group is dedicated to developing the next generation of Artificial Intelligence—systems that move beyond prediction to understand and influence the true causes behind complex processes.

Our mission is to harness causal reasoning to help and serve humanity, empowering people and organisations to make smarter, earlier, and more sustainable decisions.

The world is changing rapidly. Climate shifts, ecological risks, and social disruptions demand AI that is not only powerful but also responsible, explainable, and resilient. By uncovering root causes, identifying hidden variables, and modelling the outcomes of interventions, we create tools that allow society to go beyond “what is happening” to answer the deeper question:

What is the ideal sequence of actions—given our resources—to achieve the best possible future?

Research Focus

• Causal LLMs: redefining intelligence through the power of causality.
• Causal Agentic AI: enabling autonomous systems to reason about interventions, consequences, and long-term strategies with causal grounding.
• Causal Discovery: revealing hidden drivers behind data patterns.
• Counterfactual Reasoning: answering “What-If” questions with scientific precision.
• Robust Generalisation: building immunity to spurious correlations and distribution shifts.
• Intervention Modelling: predicting the impact of real-world actions before they are taken.
• Optimising Future Outcomes: suggesting ideal sequences of interventions under constraints such as cost, time, or resources to optiimise future outcomes.

Applications

⚡ AI for Science & Industry

• Accelerate the discovery of clean energy materials and sustainable technologies (e.g., Winner of Open Catalyst Challenge at NeurIPS AI for Science 2023, using AI to discover energy material).
• Optimise industrial operations by identifying true efficiency drivers.
• Improve manufacturing resilience by modelling intervention impacts on safety, quality, and productivity.
• Support health and sport science through causal insights into training, performance, and wellbeing.

🚚 Critical Supply Chains & Infrastructure

• Uncover the root causes of bottlenecks and disruptions in global supply chains.
• Model the cascading impacts of shocks (e.g., pandemics, conflicts, extreme weather) on trade and logistics.
• Test intervention strategies such as diversification of suppliers or rerouting of logistics.
• Support resilience planning for critical infrastructure such as energy grids, transport networks, and digital systems.

🏥 Health & Medicine

• Distinguish causal factors in disease progression, treatment effectiveness, and patient outcomes.
• Personalise treatment strategies by predicting the impact of medical interventions.
• Improve public health planning by modelling the effects of policies on populations.
• Build robust, interpretable models that doctors and policymakers can trust.

🌱 Agriculture & Food Security

• Detect crop stress and soil changes before visible damage occurs.
• Optimise fertiliser, irrigation, and planting strategies for sustainable yield.
• Model climate-driven risks such as drought, pests, and disease outbreaks.
• Support food security through resilient farming systems.

🌊 Water & Environmental Management

• Reveal hidden drivers of harmful algal blooms and water contamination.
• Model ecological responses to policy changes or land use interventions.
• Provide early warnings for floods, fires, and other natural hazards.
• Support integrated catchment management and resilient water systems.

🌍 Sustainability & Development

• Identify leverage points for emissions reduction and biodiversity protection.
• Explore sustainable development pathways under different policy or investment scenarios.
• Model land restoration strategies to optimise ecological and economic outcomes.
• Integrate satellite, climate, and policy data for systems-level planning.

🛡️ Risk, Security & Governance

• Anticipate systemic risks across finance, environment, and geopolitics.
• Provide robust evidence for policy and governance under uncertainty.
• Design interventions that reduce vulnerabilities in critical national systems.
• Support ethical and responsible AI deployment by exposing hidden biases and unintended consequences.

Impact

• Farmers increase yield while reducing environmental footprint.
• Agencies anticipate and mitigate ecological risks before crises emerge.
• Sustainability organisations balance development with stewardship of the Earth.
• Industries design smarter, safer, and more efficient processes.
• Governments build resilience into supply chains, infrastructure, and policy decisions.

By focusing on cause-and-effect, we ensure AI is not just reactive, but proactive and truly aligned with long-term human and planetary wellbeing.

Leadership

The Causal AI Group is led by Professor Javen Qinfeng Shi, a global leader in Responsible and Causal AI.

• Ranked #4 worldwide in Causation and #6 in Probabilistic Graphical Models (Google Scholar).
• Founding Director of the Causal AI Group, and Interim Director of the Responsible AI Research Centre.
• Awarded the ACM SIGIR 2025 Test of Time Award and 1st Place at the NeurIPS Open Catalyst Challenge 2023 (AI for sustainable materials).
• Proven track record of transferring research into real-world solutions across agriculture, mining, bushfire resilience, manufacturing, and more.

Group Members

Director

• Professor Javen Qinfeng Shi

Current Research Staff

• Ehsan Abbasnejad – RL, causality, VQA, agtech and sport
• Dong (Edward) Gong – continual learning, memory network, computer vision
• Zhen Zhang – causation and RL
• Yuhang Liu – latent causal representation learning
• Xinyu Li – ML-driven new energy material discovery
• Jinan Zou – causal agtech, NLP, fintech
• Ruoxin (Ritter) Wang – AI and Education

Engineering Staff

• Working on projects across agriculture, mining, sport, manufacturing, automated trades, computer vision, water utility, health, and education.

Current PhD Students

• Hossein Allahresani – Causal Vision-Language Models (VLM)
• Meisam Mahmoodi – Causal Reinforcement Learning (RL)
• Yijia Song – Causal AI for Materials Discovery
• Jiayu Huang – Causal Agentic AI
• Wenkang Jiang – Causal AI for Genomics
• Yichao Cai – Causal Representation Learning and Multimodal Integration
• Tianjiao Jiang – Causal Few-Shot Learning
• Hamed Damirchi – Task and Steering Vectors, and Low-Rank Adaptations (LoRA)

Past Staff

• Mingkui Tan – optimisation, deep learning, graphical models
• Qingsen Yan – continual learning, computer vision, crop breeding
• Xinyu Zhang – person RE-ID, medical AI
• Mahdi Kazemi Moghaddam – RL, bushfire, health
• Amin Parvaneh – ML for cave mining, active learning, RL
• Yongliang Qiao – causal effects, agtech, health

Past Students

• Bahram Mohammadi – Visual Navigation with Knowledge Reasoning
• Mohamed Khalil Jabri – RL, imitation learning, robotics
• Yuhao Lin – Segmentation and Counting
• Haiyao Cao – RL and Causality
• Hongrong Cheng – LLM compression
• Jinan Zou – NLP and Fintech
• Xian Wang – Music Transcription
• Mahdi Kazemi Moghaddam – Visual Navigation, model-based RL
• Amin Parvaneh – Active Learning, Dialogue Systems, Robotic Agents
• Xiongren Chen – Causality and Finance
• Alireza Abedin Varamin – Deep Set Learning
• Jie Yang – Reflection Removal, Signal Separation
• Suwichaya Suwanwimolkul – Compressive Sensing, Graphical Models, Healthcare
• Chongyu Liu – Visual Tracking
• Lei Zhang – Hyperspectral Image Processing
• Dong (Edward) Gong – Image Deblurring, Optimisation, Deep Learning
• Zhen Zhang – Message Passing, Higher-order Potentials
• Rui Yao – Structured Visual Tracking
• Zhenhua Wang – Inference with Unknown Graphs

Join Us

We welcome collaborations with researchers, industry partners, and policymakers who share our vision of AI that uncovers causes, not just correlations. Together, we can build technology that shapes a sustainable, resilient, and thriving future.