Bloomberg is testing ASKB, a chatbot-style AI interface for the Terminal, to help finance professionals synthesize insights from vast datasets using natural language.

Bloomberg is testing ASKB (pronounced ask-bee), a chatbot-style interface built atop multiple language models, with the beta open to roughly a third of the Terminal's 375,000 users as of publication; the company has not specified a date for full release.

ASKB lets users pose high-level investment questions in natural language—such as 'How is the war in Iran and a change in oil prices going to affect my portfolio?'—and create workflow templates to automate repetitive research tasks like comparing peer valuations, reviewing documents, and synthesizing bull and bear cases during earnings season.

Bloomberg employs validation checks at every step to minimize hallucinations (incorrect or invented information), including checks on summary facts against source paragraphs, semantic language verification, and citation validation; the company emphasizes transparency by using these systems to drive users to sources rather than hide them.