When Bloomberg dropped ASKB into the Terminal on February 23, the message to every fintech company that spent years building financial research tooling was straightforward: you are now building on Bloomberg's layer or against it.
ASKB, which stands for Ask Bloomberg, is currently in beta. The system runs a coordinated network of AI agents in parallel, dynamically retrieving, interpreting, and synthesizing data from across Bloomberg's content universe. Users ask questions in natural language; the system responds with synthesized answers grounded in Bloomberg's proprietary datasets, with attribution to underlying research documents and news sources. Where queries involve data analysis, ASKB generates the associated Bloomberg Query Language (BQL) code, which users can paste into Excel, BQuant Desktop, or BQuant Enterprise for further modeling.
The scope is large. ASKB searches across hundreds of millions of company documents, approximately 5,000 original news stories and 1.1 million curated stories published daily, and sell-side and independent research from over 800 providers. That provider count is the part that should focus the attention of anyone who has built a financial data or research business in the past decade. All of those providers are now, implicitly, components of a Bloomberg agentic workflow.
ASKB Workflows extends beyond conversational Q&A. Users can describe multi-step research tasks, such as pre-earnings preparation, post-earnings analysis, or meeting prep, and ASKB assembles structured outputs in minutes. Those workflows can be saved as reusable templates, rerun across different companies or time periods, and shared across teams.
"ASKB is a revolutionary new mode of interaction with the Bloomberg Terminal that will enable our customers to tap into the full power and context of Bloomberg's trusted data and analytics," Shawn Edwards, Bloomberg's Chief Technology Officer, said in the company's announcement. "Early feedback from beta clients shows ASKB is driving efficiency, improving discovery, and helping users surface actionable insights faster." The system is built using a combination of commercial and open-weight large language models aligned with Bloomberg's Responsible AI principles, though Bloomberg has not disclosed which specific models it uses.
David Easthope, senior analyst in market structure and technology at Crisil Coalition Greenwich, sees this as part of a broader embedding of AI into major desktop solutions. "We see these tools becoming more mainstream, and they are increasingly embedded in major desktop solutions," he told The Trade News. "As adoption accelerates, we expect AI to unlock new insights, automate complex analyses, and drive efficiency."
The adoption data, though, tells a story of wide ambition meeting cautious deployment. A Bloomberg survey of more than 300 European finance decision-makers found that 46 percent anticipate only incremental automation gains over the next three years, while 37 percent expect more far-reaching transformation. A separate Forbes-reported study found that only 6 percent of finance leaders have scaled agentic AI. According to research compiled by Neurons Lab, 66 percent of organizations already using agentic AI extensively expect to change their operating model and redefine roles, for example by flattening hierarchies and reducing middle management.
That gap between what Bloomberg is building and what finance has actually deployed is the interesting question the ASKB beta cannot yet answer. For all the talk of transformation, Bloomberg's move is as much defensive as it is offensive. The Terminal has been the dominant professional financial data platform for over thirty years, but AI-native research tools and alternative data providers have gradually eroded the moat around what professional financial research means. ASKB is Bloomberg's answer to whether a terminal designed in the 1990s can remain the central intelligence layer in an agentic world.
The limits are worth noting. ASKB is in beta; the 800-provider coverage, the BQL integration, and the Workflows templating system are features of a product not yet in general release. The LLMs powering the system are not named. The Responsible AI principles governing it are not a published technical standard. For a platform increasingly used to support investment decisions, those are open questions: not deal-breakers, but material ones.
Bloomberg has staked a claim. The 800 research providers, the BQL integration, the Workflows system: together they describe an infrastructure layer that positions ASKB as the orchestration surface for financial research. Whether that claim holds depends on whether the beta converts to production and whether the adoption gap closes. The plumbing is real. The verdict is not in yet.
