Primitive has introduced “return on agent capital” (ROAC) as a new capital class for banks, aimed at providing a standardized metric to scale AI initiatives, which currently fail 95% of the time due to a lack of justification from CFOs. This framework, launched in partnership with MX, will reach 1,700 financial institutions immediately and is designed to benchmark the performance of AI lending agents—demonstrated to deliver 3.3 times the output at just 22% of the cost. If ROAC gains traction as the industry standard for agent performance, it could redefine the landscape of regulated finance, much like how return on equity (ROE) serves as the benchmark for bank profitability.

MX: MX Technologies, Inc. is a data and software solutions provider serving financial institutions. It has entered a strategic partnership with Primitive to develop and distribute AI-native growth agents that leverage institutional data for tasks such as deposit acquisition and personalized campaigns.
Primitive: Primitive is a fintech company that has launched an AI agent operating system purpose-built for regulated financial institutions. It introduced Agent Capital as a distinct enterprise capital class and Return on Agent Capital (ROAC) as the corresponding performance framework. The company is partnering with MX to embed these tools into banking operations and workflows.

Partnership: Primitive and MX are collaborating on AI-native growth agents that integrate MX data capabilities with Primitive’s agent operating system for regulated financial institutions.
Industry Context: ROAC is positioned to serve as a benchmark for agent performance in regulated finance in a manner comparable to how ROE functions for bank profitability.
Framework Introduction: Primitive launched ROAC alongside the concept of Agent Capital to provide banks with a standardized way to measure and justify AI agent performance.