This week saw major advances in agentic AI for scientific research. A key event was the ICLR 2025 workshop in Singapore, where experts discussed challenges like improving how AI systems work together and validating their results.

Clarivate launched new AI Agents for academia, including a tool that lets schools build custom AI helpers without coding. Meanwhile, Google revealed a method where AI agents create their own training data, potentially speeding up discoveries in fields like medicine.

Researchers emphasized the need for ethical guidelines to ensure AI-driven science remains trustworthy. These tools aim to help scientists generate ideas faster while tackling complex problems like climate change and disease.