Recent research demonstrates that LLM-based agents from Anthropic and OpenAI can effectively perform phenotype annotation, traditionally a labor-intensive task for human experts. These agents showed performance comparable to trained human curators, indicating potential for scaling this process in biological research.