The Research Provenance Standard

A framework for governed AI consumption of investment research, proposed by BlueMatrix and built for the industry. AI is consuming sell-side research at scale — without attribution, governance, or any signal returning to the firms that produced it.

Persistent Attribution

Analyst identity and institution travel with the content through AI synthesis. Every AI-generated output can be traced back to its source research.
Declared Rights, Machine-Readable

Every piece of research carries a machine-readable declaration of what AI systems are permitted to do with it. Six use-type declarations. Buy side consumes programmatically with respect to rights.
AI Consumption Visibility

Every AI retrieval event logged and returned to the publisher. See which research is being queried, by which firms, and whether it shaped an investment decision.
Auditable Provenance Chain

An immutable, append-only audit trail from original research to AI-derived output. Every consumption event recorded. Every query traceable back to its source.

The window to define these defaults is open. It will not be open indefinitely.

BlueMatrix is leading the development of RPS in collaboration with a Client Advisory Board of sell-side institutions and asset managers. Join the conversation.

Talk with an expert

ai-consuming-research

AI Is Consuming Research. The Infrastructure to Govern It Does Not Exist Yet.

Buy-side firms are feeding sell-side research into AI systems right now: summarizing it, querying it, building signals from it. The analyst gets no credit. The institution gets no visibility. The compliance team has no audit trail. And the feedback loop that connects production to consumption has collapsed entirely.

This is not a hypothetical scenario. Production systems are embedded in buy-side workflows today. The infrastructure decisions being made inside buy-side technology teams right now will govern how sell-side research enters AI systems for the next decade. Those decisions are being made without sell-side input.

Five Principles. One Standard.

The Research Provenance Standard defines how sell-side investment research enters AI systems and on what terms. It is not a restriction on AI consumption. It is the framework that makes governed AI consumption possible.

RPS is built as an extension to the BlueMatrix XML schema, already in production at over 100 research departments globally. Five governance elements map directly to five core principles: Structured Delivery, Declared Rights, Persistent Attribution, Provenance Chain, and Measured Consumption.

for-sell-side-firms

For Sell-Side Firms: Define the Terms. See the Impact.

Sell-side firms already have disclaimer language prohibiting AI training and fine-tuning without consent. The problem: prose disclaimers cannot be parsed or enforced by AI systems. RPS encodes those same positions in machine-readable form that AI systems can actually check.

With RPS, sell-side firms can configure AI rights per document and per content type, manage entitlement requests through a governed queue, and receive AI consumption analytics for the first time. Not just whether research was downloaded, but whether it shaped a trade idea or portfolio decision.

for-buy-side-firms

For Buy-Side Firms: Attribution You Can Trust. Signals You Can Act On.

Buy-side firms using AI to synthesize research face the same visibility problem in reverse. When an AI system generates an answer, there is no record of which broker's research shaped it. No attribution, no audit trail, and no way to demonstrate the value of sell-side relationships to investment leadership.

RPS changes that. Every AI-generated output is traceable back to its source research. Broker influence scoring shows which relationships are actually driving alpha, to allocate research budgets strategically. And broker coverage gap analysis gives portfolio teams and investment leaders the data to identify opportunities to get the most value out of broker relationships.

built-on-27-years

Built on 27 Years of Structured Research Infrastructure.

The Research Provenance Standard is not a proposal from the outside. It is the next layer of infrastructure from the inside, built by the platform that already powers how research is created, governed, and distributed across the industry.

BlueMatrix has spent 27 years building the XML schema that structured research runs on. That schema is already in production at over 1,000 research departments globally. RPS is a fully additive extension of that schema. No existing documents are modified or deprecated. Everything already running continues to run.

-->