S&P Global launches dual data retrieval for AI agents

S&P Global (NYSE: SPGI) announced the launch of Adaptive Retrieval, a new service enabling AI agents and large language models to access and assemble S&P Global's licensed data using natural language queries. The service launches alongside the company's existing Deterministic Retrieval method, both now available through the S&P Global AI Data Portal, creating what the company describes as the first offering to provide two complementary data retrieval methods together.

Adaptive Retrieval is designed to pull data from multiple sources simultaneously and automatically handle requests spanning multiple datasets, making it well-suited for complex, multi-step tasks including in-depth research and report generation. The Deterministic Retrieval method, built on the Kensho LLM-ready API and available to customers since 2025, provides API-driven access to S&P Global data through direct, structured queries and is ideal for focused tasks such as researching a specific company or analyzing earnings call transcripts. Customers can deploy one method or both, depending on their system architecture and objectives.

Sally Moore, Chief Client Officer and Co-Head of Market Intelligence, framed the launch within a broader industry transition: "The use of AI in financial services is rapidly accelerating and evolving, from tightly controlled workflows to fully autonomous, multi-agent systems. With Deterministic and Adaptive Retrieval now available together, we're ensuring that S&P Global's trusted data is accessible across that full range of workflows, so customers can access data the way they need it today and adapt as their architectures evolve."

The underlying rationale centers on how AI systems now operate in enterprises. As organizations transition from human-driven processes to AI-driven workflows where agents execute tasks autonomously, data requirements have shifted fundamentally. AI agents and LLMs require data that is properly cited, verifiable, and auditable—qualities that historically demanded significant engineering work to achieve: identifying and validating sources and building custom retrieval logic requiring deep domain expertise. By pre-structuring S&P Global's data and making it available through the portal, the company aims to eliminate that complexity, allowing customers to focus on building products and generating insights rather than preparing and managing data. Bhavesh Dayalji, Head of Kensho Data & Intelligence, indicated that this retrieval layer is foundational to broader work ahead: "Now, S&P Global data flows directly into the tools and platforms where customers work through financial skills and plugins, and MCP apps that allow customers to visualize, explore, and interact with S&P Global data inside AI applications."

S&P Global's announcement reflects a fundamental shift in how financial services organizations are adopting AI. The body identifies the key transition: from tightly controlled workflows where humans direct tasks to fully autonomous, multi-agent systems where AI agents operate independently. This transition creates new demands for enterprise data—it must be properly cited, verifiable, and auditable in ways traditional data integration was not designed to guarantee.

Historically, integrating high-quality data into AI systems required significant engineering effort: finding and validating sources, building retrieval logic, and maintaining domain expertise throughout. By offering both Deterministic and Adaptive Retrieval through a single portal, S&P Global aims to eliminate that complexity by providing data that is already structured, cited, and audit-ready. This positions the portal not as a passive data store but as a foundation layer for what the company calls "AI-native experiences"—tools and workflows designed around how AI systems actually work rather than how humans traditionally access data.

No comments yet. Be the first to share your thoughts!