JASA Special Issue Explores Role
of Statistical Science in AI
The influence of artificial intelligence continues to grow in research, industry, and daily life. In response, the Journal of the American Statistical Association has announced a special issue, “Statistical Science in AI.”
With AI rapidly reshaping our world, the statistical community stands at a crossroads. In planning this issue, the JASA editorial team—led by Jianqing Fan, Annie Qu, and Dylan Small, along with guest editors Xihong Lin and Susan Murphy—identified the following three critical areas in which the statistical community can make the greatest impact:
- Identification of core AI problems: Highlighting contemporary AI challenges in which statistics can play a vital role in modern study designs, federated learning, generative AI, and critical issues of interoperability and ethics
- Statistical contributions to AI: Demonstrating how statistical theories offer unique solutions to AI bottlenecks, empowering these tools to advance real-world science beyond simple pattern recognition
- Innovative theory and application: Developing new models and inventive analysis methods that move beyond existing techniques, specifically those aimed at solving the practical problems addressed by AI
As guest editor, Lin observes, “AI is transforming science and society. It presents extraordinary opportunities, but also profound methodological and real-world challenges. This special issue showcases research that brings together statistical thinking and principles at the forefront of AI, ensuring rigorous uncertainty quantification, and advancing trustworthy and transparent ‘white-box’ methods.”
According to Hongtu Zhu, editor of JASA: Applications and Case Studies, the introduction of a LArge Model Based Data Agent is more than just a new technical tool. It signals a fundamental evolution in the discipline’s future, with major shifts in how statistical work will be conducted, taught, and evaluated. Rather than a traditional standalone method, LAMBDA frames data analysis as an agentic workflow designed to improve reliability and scalability. In the private sector, heavy investment in AI agents is rapidly reshaping data analysis, with clear implications for statistics education, research, and workforce development. Zhu states, “I view this as closely aligned with the field’s future, even if it doesn’t fit a traditional template. It highlights emerging forms of applied statistical practice and the new methodological and theoretical challenges it creates.”
This special issue maps evolving territory, where statistical theory meets core AI challenges. From deep learning and reinforcement learning to ethical puzzles of data privacy, JASA aims to highlight research that pioneers “trustworthy AI”—systems that are powerful, but also interpretable, scalable, and scientifically sound.
“We believe sound statistical foundations and deeper integration between AI and statistics will advance both fields and empower real-world scientific discovery,” Lin adds. “We hope this collection inspires the next generation of principled methodological development and impactful applications, fostering deeper dialogue and engagement across disciplines.”
As the ASA continues to lead the conversation on the role of data in society, this is an opportunity for the community to shape this discussion at a critical point in time. By integrating statistical rigor with AI’s computational power, researchers are not just improving algorithms—they are ensuring the future of technology is built on a foundation of reliability and scientific truth.
With a future increasingly defined by automated decision-making, the perspective of the statistical scientist will be more vital than ever. JASA’s dedication to this intersection ensures our field remains at the heart of the AI conversation.
Visit the ASA Member Hub for details on how to access JASA and other ASA journals.