A Guth Labs publication ↗
Back to News

Performativ opens governed portfolio-data access to AI assistants

Blank cream record folders inside a green glass vault secured by a gold gate and key.
AI-generated editorial illustration

Performativ launched an updated Model Context Protocol connection on November 19 that lets staff at banks and wealth managers query client and portfolio information through assistants including ChatGPT and Claude.

The company announcement says access is available to existing Performativ clients. It is disabled until a firm enables it, and administrators decide which employees can connect.

Each user signs in with their own account and retains existing Performativ permissions. Requests are logged, and Performativ hosts the connection in the European Union, according to the release.

Teams can ask about holdings, performance, risk, regulatory profiles, meeting notes and recent account activity. Suggested workflows include preparing client-review briefings and reviewing changes to client records.

The update builds on Custom Agents capabilities introduced in November 2025. Direct access for a bank's own clients through their chosen assistants remains a future ambition, not an announced generally available service.

The practical change is bringing existing financial workflows into outside AI interfaces while retaining firm-controlled access. The announcement describes governance features, not an independent security audit or a blanket guarantee of regulatory compliance.

Sources and citations

Revision history

Published .