Artificial Intelligence expands across scientific research

Artificial Intelligence is taking a larger role across biology, chemistry, physics, astronomy, and earth science, with publication volume rising sharply and new scientific infrastructure emerging. Performance gains are notable in narrow tasks, but current systems still struggle to replicate research and complete end-to-end scientific work at expert level.

Artificial Intelligence is becoming more deeply embedded across scientific domains including biology, chemistry, physics and astronomy. Natural sciences reached approximately 80,150 Artificial Intelligence publications in 2025, up 26% from 2024. Artificial Intelligence now accounts for 5.8%-8.8% of scientific research output depending on the field, up from below 1% in 2010.

Performance is advancing in several specialized areas, but reliability remains uneven. On ChemBench, the best models surpass human expert averages across 2,700+ chemistry questions while struggling with basic tasks. On ReplicationBench, frontier models score below 20% on paper-scale replication in astrophysics. On UnivEarth, LLM agents answer earth observation questions with 33% accuracy, and their code fails 58% of the time. On end-to-end scientific research tasks, the best Artificial Intelligence agents score roughly half of what PhD experts achieve. On PaperArena, the best agent reaches 38.8% accuracy versus a PhD expert baseline of 83.5%. On BixBench, frontier models achieve roughly 17% accuracy on real-world bioinformatics analysis.

Scientific infrastructure is also shifting toward larger Artificial Intelligence-native systems and datasets. Astronomy released its first foundation model, first visualization benchmark, and a 100TB training dataset in 2025, signaling a field-wide shift toward Artificial Intelligence infrastructure. AION-1, trained on over 200 million celestial objects from 5 major surveys, is the first astronomy foundation model. AstroVisBench introduced the first benchmark for LLM scientific computing and visualization in the field.

Weather and climate research saw a major operational step forward. An Artificial Intelligence system ran a full weather forecasting pipeline end-to-end for the first time in 2025. Aardvark Weather replaced the traditional numerical prediction pipeline with a single ML system, and multiple Artificial Intelligence weather models reached operational deployment. FourCastNet 3 generates a 60-day global forecast in under 4 minutes, running 8 to 60 times faster than prior approaches.

Research automation also moved forward, though confirmed scientific impact remains limited. The first fully Artificial Intelligence-generated paper was accepted at a peer-reviewed workshop in 2025, but the list of experimentally confirmed Artificial Intelligence discoveries remains short. Sakana’s Artificial Intelligence Scientist-v2 produced a paper accepted at an ICLR workshop without human-coded templates. Google’s Artificial Intelligence Co-Scientist was validated in three biomedical areas. Most Artificial Intelligence models for science still come from academic and government institutions, while industry leads foundation model development in weather and climate.

58

Impact Score

GCC accelerates Artificial Intelligence strategy

Gulf states are embedding Artificial Intelligence into national economic plans, pairing state-backed investment with new governance frameworks and digital infrastructure projects. The region is positioning itself as a sovereign Artificial Intelligence hub spanning data centers, cloud capacity, and sector-specific deployment.

YMTC expands memory production with new fabs

YMTC is preparing a major manufacturing expansion that would more than double its wafer output and extend its push beyond NAND into DRAM. The company is also increasing its reliance on domestic equipment as trade restrictions continue to shape its supply chain.

Ex Parte Desjardins reshapes Artificial Intelligence patent eligibility

A precedential PTAB decision and two USPTO memoranda have clarified how Artificial Intelligence inventions can qualify for patent protection under 35 U.S.C. § 101. The guidance gives applicants a clearer path to showing technical improvements in machine learning systems and computer performance.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.