Bloomberg has quietly started to roll out its own version of ChatGPT for the terminal used by 325,000 financial professionals.

The function, ASKB (i.e. Ask Bloomberg), lets you ask questions in plain English about research, market prices, news and other data.

It’s notable how long it took to roll out (three years after generative AI); how limited it remains (you can’t ask about a portfolio); and how it’s not fully integrated into the design (i.e. it’s a separate window, like Bloomberg Chat.) Competitors FactSet and AlphaSense have already launched similar solutions.

Still, it marks an enormous step forward and has the potential to change the workflow for how Wall Street professionals operate day to day.

Small batches of clients are being enabled to gather feedback and work out bugs. A few clients have tweeted about the function, saying it looks promising but has shortcomings.

Chuck Trafton, CIO at CLH Partners, tweeted that ASKB has strengths like natural language queries, access to market prices, source transparency and fewer hallucinations. Drawbacks include it can’t access portfolios and market data is not real time.

Alex Thorn, head of firmwide research at Galaxy, tweeted that it got hung up for two hours when he asked: “Analyze the constituents of MSCI funds and tell me a list of companies that would be considered companies that are “non operating businesses:”

Parker Evans, CFA, CFP, Founder of Successful Portfolios,  tweeted ASKB is getting better. His question was: “What are the implications for investors of the election results in Argentina?”

Matthew Sigel, recovering CFA, head of digital assets at Van Eck, tweeted this question: “Is there any evidence that companies adhering to “rule of 40” framework have stock prices that outperform?”

I haven’t spoken to Bloomberg about the feature, so I don’t know if ASKB is leveraging an existing LLM like OpenAI, Claude, or Gemini or a home-built Bloomberg model.

Connecting directly to Bloomberg data should make hallucinations less likely.

On the other hand, it’s not connected to the open Internet, so it won’t be able to answer a lot of questions everyone has come to expect from LLMs.

There is a slew of startups that leverage LLMs to tap into financial insights, including one I advise, DoTadda, Inc.. Many of those firms currently provide faster and better AI insight features.

Bloomberg has a deep moat that means it doesn’t have to rush. It will undoubtedly continue to innovate.

One unknown is whether Bloomberg will eventually charge more to use an AI tool like ASKB, something I’ve been told FactSet does.

Financial data terminals are already expensive, adding a fee to access the data via a chat bot on top of that could dramatically increase costs.