Press releases

‍

New natural-language agents designed to accelerate discovery and surface deeper insights

COLUMBUS, OH – October 6, 2025 – CAS, a division of the American Chemical Society specializing in scientific knowledge management, today announced the pending launch of science-smart agentic AI experiences in CAS BioFinder® and CAS SciFinder®. These capabilities introduce intelligent agents trained on scientific content to automate complex research workflows. Scientists can ask direct questions and receive reliable, evidence-backed responses grounded in the curated data of the CAS Content Collection™, unlike generic AI tools that rely on unverified or incomplete sources.

Expanding agentic AI across CAS platforms and services

In the upcoming months, CAS BioFinder users will gain access to a conversational AI agent that allows researchers to pose complex scientific questions and follow up on responses. To support CAS BioFinder’s global reach, users can interact with the agent and receive detailed responses in their preferred language. A literature review agent is also planned, designed to summarize relevant publications and support broader exploration. CAS SciFinder will introduce similar agentic AI capabilities before the end of 2025, expanding support for chemical research workflows.

CAS has long been a leader in scientific rigor and transparency, and this agentic AI experience maintains those standards. Responses are linked to CAS entities and external sources for validation and further exploration. Judgment agents flag potential inaccuracies, and users can resume or download session histories for review. These capabilities are developed in alignment with the CAS Ethical Approach to AI, which emphasizes the importance of pairing technology with scientific expertise to support responsible research in ways that reflect the values of the global research community.

These advancements reflect the broader commitment of CAS to prepare researchers for the future of scientific discovery. Agentic AI, trained on scientific language, workflows, and the CAS Content Collection™, is being integrated across CAS solutions to enhance scientific workflows, within internal operations to improve efficiency and accuracy, and through CAS Custom Services℠ to help customers advance their own AI strategies.  

Why it matters

Traditional application interfaces support structured exploration and reliable results. The agentic AI experience builds on that foundation by enabling scientists to ask research questions in natural language and receive accurate, evidence-based responses. Designed to enhance expertise, not replace it, this approach offers a more intuitive way to navigate scientific information while preserving the flexibility of existing workflows.

By drawing on the CAS knowledge graph, agentic AI connects concepts, relationships, and data points that may not be immediately visible through conventional searches. This deeper contextual understanding helps researchers uncover insights faster and make more confident decisions.

Agentic AI in CAS BioFinder is helping researchers explore complex disease areas. In one case, users investigating ovarian cancer engaged with a conversational agent to guide inquiry and uncover new directions. While these insights could be reached using existing CAS BioFinder capabilities, the agentic experience offers a new way to interact with scientific content, allowing researchers to ask natural-language questions and receive structured, report-style responses that streamline exploration.

“Agentic AI saves scientists time and helps them focus on higher-value work,” said Manuel Guzman, President, CAS. “These agents accelerate the innovation process by streamlining data exploration, highlighting relevant insights, and enabling faster, evidence-based decisions.”

Building more reliable AI for diverse research challenges

CAS is committed to advancing agentic AI capabilities to support reliable research acceleration in diverse environments. Future development will focus on expanding agent functionality to address complex challenges in predictive modeling and decision support.

These capabilities will enable organizations to uncover deeper insights from the human-curated data in the CAS Content Collection, refine their research strategies, and accelerate innovation across disciplines.

About CAS

CAS connects the world's scientific knowledge to accelerate breakthroughs that improve lives. We empower global innovators to efficiently navigate today’s complex data landscape and make confident decisions in each phase of the innovation journey. As a specialist in scientific knowledge management, our team builds the largest authoritative collection of human-curated scientific data in the world and provides essential information solutions, services, and expertise. Scientists, patent professionals, and business leaders across industries rely on CAS to help them uncover opportunities, mitigate risks, and unlock shared knowledge so they can get from inspiration to innovation faster. CAS is a division of the American Chemical Society. Connect with us at cas.org.

‍