Researchers at the NIH developed GeneAgent, an AI-driven large language model system that autonomously verifies gene-set functional analysis predictions against expert-curated biological databases, significantly reducing inaccurate assertions common in prior LLM tools. GeneAgent applies a self-verifying approach to cross-check initial outputs, enhancing the reliability of gene-set enrichment and functional descriptions in genomics. This advancement offers a substantial improvement for computational biology and bioinformatics research.