Intercontinental Exchange (ICE) (NYSE:ICE) has taken a significant step toward integrating its proprietary fixed income information directly into modern artificial intelligence environments. On July 29, 2026, the company announced that licensed clients can now access key fixed income datasets and related methodologies through major AI platforms via the Model Context Protocol, commonly known as MCP.
This development aims to create an intelligent knowledge layer that sits atop ICE’s established data and analytics offerings.
Users interacting with their preferred AI tools can receive answers grounded in ICE’s specialized content rather than generic model knowledge alone.
The approach combines structured market information with the analytical capabilities of large language models, allowing professionals to query complex fixed income topics in a conversational manner while remaining within a controlled, permissioned setting.
MCP originated with Anthropic and functions as an open standard overseen by the Agentic AI Foundation.
It enables secure connections between large language models and enterprise data systems.
Compatibility extends across leading services such as Claude, ChatGPT, Gemini, and Microsoft Copilot.
ICE’s implementation currently operates within Claude, with the open nature of the protocol designed to support broader platform availability over time.
A protected knowledge layer forms a core element of the offering.
This layer preserves the integrity of ICE’s proprietary methodologies, data hierarchies, and interrelationships.
As a result, the underlying intellectual property remains safeguarded even as clients incorporate the information into AI-driven processes.
Chris Edmonds, President of ICE’s Fixed Income and Data Services, highlighted the significance of the move.
He noted that customers can now embed the firm’s industry-leading data, methodologies, and domain expertise into their AI workflows for the first time.
The model pairs this specialized content with AI reasoning capabilities. Importantly, the structure emphasizes payment based on the value of information delivered rather than the volume of computational resources consumed.
This creates a more predictable and transparent cost framework for incorporating ICE data into decision-making processes.
Several specific fixed income datasets are included in the initial release.
End-of-day fixed income evaluations cover more than three million instruments spanning over 150 countries and more than 80 currencies.
Enhanced Evaluation Transparency supplies roughly 160 data points per security identifier, encompassing basic reference details, commentary on price movements, mortgage-backed securities assumptions, and trade or quote information.
Additional resources include the ICE AAA Municipal Bond Curve, a transaction-driven tool used both as a standalone product and within ICE’s evaluation processes; US Treasury Benchmark Data, similarly available on a standalone basis and integrated into evaluations; and trade data from FINRA TRACE covering corporate, agency, and securitized products alongside municipal bond trades reported through MSRB RTRS.
By delivering these resources through an open connectivity standard, ICE aims to support more seamless incorporation of high-quality fixed income information into the AI tools that market participants increasingly rely upon. The focus on value-based access and maintained data integrity positions the initiative as a practical bridge between traditional financial data infrastructure and emerging generative AI applications.