In their battle for enterprise sales, both OpenAI and Anthropic have been targeting financial services firms. OpenAI has a team of ex-investment analysts building a yet-to-be-launched agentic AI financial analysis product. Anthropic has been rolling out financial modeling skills for its Claude Code, Coworker, and Claude for Finance offerings. Startup Samaya AI is building AI tools for the finance sector as well.
The veteran of specialized financial data and analysis tools is Bloomberg. Access to the company’s terminal remains the standard tool for traders, investment bankers, and hedge fund quants.
But today, AI is intensifying competitive pressure on Bloomberg as rivals embrace AI-powered features and use AI models to rapidly analyze complex data that once only Bloomberg consolidated in a single place.
Bloomberg recently unveiled “AskB,” which it calls the biggest rethink of the terminal in the company’s history. AskB allows users to navigate the terminal’s features using natural language, but it does far more. The system acts as an agent, building investment screens and producing full research reports including sophisticated financial modeling and bull and bear cases for particular stocks on demand.
AskB uses a variety of AI models under the hood, including some built by Bloomberg itself and others from frontier AI model companies such as Anthropic. Shawn Edwards, Bloomberg’s chief technology officer, explained how the company built AskB and what enterprises in any industry can learn about getting agentic AI to deliver real business value.
Data remains the critical differentiator
AskB pulls from Bloomberg News, sell-side research from over 800 providers, market data, and increasingly so-called alternative datasets that are hard or expensive to source. This includes anonymized credit card transactions, foot traffic in retail locations from cellphone pings, satellite imagery of parking lots, and app usage data.
Having this data in one place allows the AskB agent to perform powerful analyses, such as aligning data with the business segments a public company reports in order to “nowcast” quarterly KPIs. Before Sweetgreen’s fourth-quarter 2025 earnings call, alternative data signaled the chain would miss analysts’ consensus earnings forecasts — which it ultimately did.
When asked whether customers could simply use AI models to ingest this data and run analyses themselves, obviating the need for Bloomberg’s approximately $30,000-per-user annual subscription, Edwards said a few have tried and found it harder than it looks.
“You have to buy all those sources, do all the validation work, build benchmarks — and tokens aren’t cheap. Most customers are saying, ‘Awesome, Bloomberg, you do that. I’m going to focus on my [own trading strategies],'” Edwards said.
AI has also dramatically accelerated how Bloomberg builds datasets. Data ingestion that used to take four-and-a-half months now takes two days, freeing up teams once dedicated to data entry and cleaning to focus on building internal evaluations.
Building robust evaluations is essential
Building good internal evaluations is critical to deriving ROI from AI agents. “Evaluations, I cannot stress enough, are the make-or-break of building a useful, trustworthy system,” Edwards said, calling the emphasis on evaluations one of the biggest cultural shifts Bloomberg has experienced in the past two years.
Building the evaluations requires close collaboration with domain specialists — bond covenant experts, equity analysts, market structure specialists, and even Bloomberg’s journalists — along with engineering and product teams. Bloomberg pulled these experts off their day jobs both to write benchmarks for sub-agents and to help evaluate entire workflows.
Using AI models as evaluators can work for straightforward cases, but for everything else, human assessors are required. Through building these evaluations, Bloomberg is encoding its experts’ tacit knowledge into how its AI agents work.
Multi-model workflows help contain costs
Cost discipline is fundamental. Workflows need to be multi-model. AskB uses a mix of commercial frontier models and open-weight ones, as well as Bloomberg’s own internal models, routing queries to the cheapest model that can handle a given task with the required reliability and performance.
The next frontier: proactive agent-to-agent workflows
Edwards’ vision for what is coming centers on agent-to-agent workflows and always-on data monitoring. He wants Bloomberg to be “the eyes and ears” for its financial customers — watching the world against each client’s positions, mandate, and strategy, and flagging not just obvious developments but second- and third-order effects.
A flood takes out a factory making parts for a supplier to a company whose stock you hold; AskB, in Edwards’ vision, would flag the problem before you had thought to ask.
Achieving that vision will be difficult, but Bloomberg is demonstrating key steps along the path.