Open and sustainable AI: challenges, opportunities and the road ahead in the life sciences.

Open and sustainable AI in the life sciences can be advanced through practical recommendations aligned with over 300 ecosystem components, addressing trust, reusability, and environmental impact.

• Why it matters: Addressing challenges in AI's reproducibility and environmental sustainability is crucial to ensure reliable, efficient, and responsible use of AI in biological research, filling a critical gap in current practices.
• What they did: The authors reviewed AI ecosystem fragmentation and proposed over 300 targeted recommendations and implementation pathways to promote open, reusable, and sustainable AI models in life sciences.
• The result: This framework facilitates resource connection, supports policy development, and guides AI deployment, enabling more trustworthy and environmentally conscious AI applications in biological research.