Kensho

Learning Objectives

• Describe what Kensho does and how it connects AI agents to authoritative financial data
• Explain the Model Context Protocol (MCP) and why governed data access matters for finance
• Recognize that Kensho is data infrastructure, not investment advice, and depends on underlying data licenses

What Is Kensho?

Kensho is the artificial-intelligence arm of S&P Global, one of the largest providers of financial data, ratings, and analytics. Its role is to make S&P Global's vast, authoritative datasets usable by AI systems. Through an "LLM-ready" API and a Model Context Protocol (MCP) server, Kensho lets large language models and AI agents query datasets such as S&P Capital IQ financials, estimates, earnings-call transcripts, market data, business relationships, mergers-and-acquisitions transactions, and private-company data — all in natural language. The problem it solves is that finance professionals and the AI tools they use need trustworthy, source-of-record data; without a governed connection, an AI assistant is left guessing or pulling from unreliable sources. Kensho provides the pipe that lets an agent retrieve verified figures instead.

Kensho operates as a subsidiary of S&P Global, which is publicly traded on the New York Stock Exchange under the ticker SPGI. Kensho ships prebuilt "financial skills" — such as tearsheets, deal digests, and earnings previews — and has announced integrations that bring S&P Global data into AI platforms including Claude, ChatGPT, and Databricks through its MCP server. This positions Kensho as connective infrastructure between authoritative financial data and the growing ecosystem of AI applications.

Core Capabilities

LLM-ready API and MCP server

Kensho exposes S&P Global data through an API designed for large language models and an MCP server, so AI applications can retrieve financial data in natural language through a standard, permissioned interface rather than bespoke integrations.

Access to authoritative datasets

The service reaches core S&P Global data, including S&P Capital IQ financials and estimates, earnings-call transcripts, market data, business relationships, transactions, and private-company financials — the kind of source-of-record data finance professionals rely on.

Prebuilt financial skills

Kensho ships ready-made "skills" such as tearsheets, deal digests, and earnings previews, giving teams higher-level building blocks so agents can assemble common finance deliverables without building each retrieval from scratch.

Integrations with AI platforms

Through its MCP server, Kensho connects S&P Global data into major AI environments such as Claude, ChatGPT, and Databricks, letting firms bring authoritative data into the tools their teams already use.

Strengths

• Authoritative data source — draws on S&P Global's source-of-record datasets, a level of trust generic web data cannot match.
• Standard, governed access — the MCP server offers a permissioned, auditable channel that respects data entitlements.
• Prebuilt skills — tearsheets, deal digests, and earnings previews speed up common finance tasks.
• Broad integrations — connects into Claude, ChatGPT, and Databricks so teams use existing AI tools with trusted data.

Limitations and Considerations

• Quality depends on data licenses. Access to S&P Global datasets is governed by licensing and entitlements; a firm only gets the data it is licensed for, and coverage and cost scale with those licenses.
• It is infrastructure, not advice. Kensho supplies data and retrieval skills, not investment recommendations. The interpretation of the data and any resulting decisions remain the responsibility of the professional and the firm.
• AI-agent limits still apply. An agent querying Kensho can still misinterpret or mis-frame authoritative data, so outputs that inform decisions warrant human review even when the underlying figures are trustworthy.
• Enterprise scope. The service targets financial-services organizations building AI tools and assumes an enterprise data relationship with S&P Global rather than casual, individual use.

Best Use Cases

Getting Started

1. Confirm which S&P Global datasets your firm licenses, since access through Kensho follows those entitlements.
2. Connect the Kensho LLM-ready API or MCP server to the AI application or agent you want to ground in authoritative data.
3. Use the prebuilt financial skills — such as tearsheets and earnings previews — to accelerate common deliverables.
4. Keep human review over any decision-relevant output, treating Kensho as trusted data infrastructure rather than advice.

Key Takeaways

• Kensho is S&P Global's AI arm, providing an LLM-ready API and MCP access to authoritative financial data.
• It lets AI agents query S&P Capital IQ financials, estimates, transcripts, and more in natural language, with prebuilt financial skills.
• Its value depends on the underlying data licenses, and it is infrastructure for retrieving data, not investment advice.
• It is best suited to financial-services teams building AI tools that need governed access to source-of-record market data.