Executive Summary
Financial institution BNY has launched "Eliza," a centralized internal platform enabling its employees to build and deploy AI agents using OpenAI's frontier models. The initiative, guided by the mantra "AI for everyone, everywhere," emphasizes a "governance by design" approach to safely scale AI use across the enterprise. With over 125 live use cases and 20,000 employees actively building agents, BNY is transforming core business operations, from legal contract review to client relationship management.
Key Takeaways
* Initiative Name: BNY's internal AI deployment and education platform is called "Eliza."
* Primary Function: Eliza enables BNY employees to safely build, deploy, and share AI agents for various business tasks within a governed environment.
* Scale of Adoption: The platform supports over 125 AI tools in production, with 20,000 employees actively building agents. 99% of the workforce has been trained on generative AI.
* Embedded Governance: A core feature is "governance by design," which integrates security, permissions, and oversight directly into the platform. This is managed by cross-functional groups including a Data Use Review Board and an AI Release Board.
* Example Use Cases:
* Contract Review Assistant: Reduces legal review time for vendor agreements by 75%.
* Risk Insights Agent: Surfaces emerging risk signals from deep research to help analysts act proactively.
* Digital Employees: Advanced agents with dedicated workflows that handle tasks like payment instruction validation.
* Advanced Capabilities: A select group is using ChatGPT Enterprise for its "deep research" capability, which enables multi-step reasoning for complex tasks like risk modeling and strategic planning.
Strategic Importance
This implementation serves as a significant proof point for OpenAI's enterprise value, demonstrating how generative AI can be securely deployed at scale within a highly regulated, systemically important financial institution. It provides a blueprint for other large organizations on how to balance widespread AI adoption and innovation with stringent risk management and governance.