In the ever-evolving landscape of technology, the financial industry is undergoing a quiet revolution, driven by the integration of AI agents and the Model Context Protocol (MCP). At QCon London 2026, Morgan Stanley's Jim Gough and Andreea Niculcea shed light on their journey of adapting their API program to this new era. Their story is a testament to the power of innovation, the challenges of scaling, and the importance of codified control in the age of AI.
The Shift to Natural Language
One of the most intriguing aspects of this transformation is the shift in how business users interact with their data. Traditionally, OpenAPI specs were met with indifference, but the introduction of MCP specs has sparked excitement. The demand for natural language interactions with data, such as trades, risk, and portfolio positions, is driving this change. However, as Gough points out, the simplicity of a handful of tools quickly breaks down when scaling to dozens, leading to disambiguation problems and increased costs.
The Role of CALM
To manage this complexity, Morgan Stanley turned to CALM (Common Architecture Language Model), an open-source project under FINOS. CALM allows teams to define architectures as code, providing a single source of truth for the intended state of a system. This approach enables the generation of everything needed for deployment from a single schema, acting as organizational templates. The result is a significant reduction in time to production, from roughly two years to just one or two weeks.
The Impact of Codified Control
The use of CALM has also had a profound impact on the culture among developers and architects. While developers may lose some flexibility in how they wire things together, they gain a working production baseline from day one that already passes every gate. This approach also allows for the bootstrapping of entire projects, pulling in frameworks like Spring Initializr or Quarkus Start, preconfigured with the right defaults.
The Future of APIs
Looking ahead, Gough suggests that the adapter layer will continue to shift, with APIs remaining the stable contract underneath. The ability to swap the interaction layer without rebuilding everything is a key advantage of this approach. As the industry continues to evolve, the importance of codified control and the integration of AI agents will only grow, shaping the future of financial technology.
In conclusion, the story of Morgan Stanley's API program is a fascinating exploration of the challenges and opportunities presented by the integration of AI agents and the Model Context Protocol. It is a reminder that innovation requires a balance between flexibility and control, and that codified control can be a powerful tool in the age of AI.
