wsj.com web signal

Bloomberg Opens ASKB AI to All Terminal Users With Custom Docs

TL;DR

  • Bloomberg is extending ASKB, its multi-agent conversational AI, from a beta subset to all users of its 375,000-seat Terminal.
  • Subscribers will be able to upload their own documents into Bloomberg's data environment, where client material is kept segregated from Terminal content.
  • Each ASKB answer cites its source and returns the underlying BQL code behind any data analysis, aimed at regulated capital-markets workflows.

Bloomberg is extending ASKB, its conversational AI for the Terminal, from a beta subset to its entire subscriber base, and will let subscribers plug their own documents and data into the system alongside Bloomberg's own content. Steven Rosenbush reported the move in WSJ CIO Journal.

About a third of the Terminal's 375,000 users had ASKB access by April 2026, and Bloomberg has not specified a date for the full release.

ASKB is, in the company's description, "a conversational AI interface for the Terminal built on specialized agents rather than one massive model," coordinating agents that search documents, news, analytics and alternative datasets in parallel. When an answer leans on data analysis, the system returns the underlying Bloomberg Query Language code so users can lift it into Microsoft Excel, BQuant Desktop, or BQuant Enterprise.

The custom-content integration is the bigger reach. Subscribers can upload files into Bloomberg's data environment, where, the company says, "client material is securely segregated from the Terminal's existing data." Bloomberg has pointed to recurring tasks such as "Customized Morning Briefs or weekly thesis health checks" as early automation candidates.

Each answer cites its original source and timestamp, which Bloomberg argues keeps the tool usable in regulated capital-markets workflows. The rollout lands in an active quarter of enterprise AI launches on our tracker, alongside this week's Honeywell Forge Cognition and Modal Series D stories.