Recently, the intelligent breeding research group from the maize team at Huazhong Agricultural University (HZAU) published a paper in the journal Molecular Plant titled "HAIant(智然体): A human-centered AI framework for secure, personalized intelligence augmentation in biological research".

The study systematically analyzes critical challenges regarding data security, privacy preservation, and domain capability enhancement, constructing a locally deployed human-centered intelligence augmentation framework alongside BioSkills bioinformatics tool modules to enable user-controlled research assistance.

Over the past decade, artificial intelligence (AI) has profoundly reshaped scientific paradigms, deeply penetrating genomics, protein structure prediction, and drug discovery while giving rise to milestone tools like AlphaFold3 and genome language models.

Large language models have further automated research workflows. However, open AI applications have sparked widespread concerns regarding data security, privacy, and cognitive impacts. Furthermore, existing systems predominantly focus on cloud-based general tasks, neglecting localized research requirements.

To address these challenges, the research proposes the HAIant framework. By establishing personal knowledge bases, integrating specialized tools, and deploying local large language models, the framework achieves continuous context awareness and task execution.

Demonstrated through Bio-HAIant, the approach streamlines experimental record management, bioinformatics analysis, and knowledge communication, providing a novel perspective for fusing AI with biological research. Ultimately, the system serves as a collaborative long-term partner rather than a replacement for human judgment.

HAIant Secure Localization Architecture. [Photo/news.hzau.edu.cn]