Announcements

WatersTechnology | Bloomberg Debuts GenAI News Summaries

January 15, 2025

This article was originally published in WatersTechnology.

By Max Bowie

The AI-generated summaries will allow financial professionals to consume more data, faster, officials say. 

Bloomberg is adding summaries created using generative artificial intelligence to its proprietary news stories on its terminal to provide key takeaways from stories and save terminal users’ time.

The three-bullet-point AI-Powered News Summaries will go live on news story written by Bloomberg journalists, starting at 7:30am New York time today, Jan. 15.

Adela Quinones, global head of news discovery at Bloomberg, says the move is part of her group’s efforts to ensure that, of the 1.5 million news stories carried on the Bloomberg terminal every day, the right stories reach the right users.

“There’s a world of content and context that clients need to consume, and it’s constantly growing. So, we’re thinking about how to make that easier to consume,” she says.

The summaries are designed to serve a range of use cases but essentially allow readers to see the key information in a story quickly and succinctly, and decide whether to read the entire article, or to collect key data points to quickly obtain an overview of a company or market sector.

“For a salesperson, it might be about sharing an interesting story with a client,” driving greater client engagement, Quinones says. “Or an analyst might be looking at a large swath of information… and without having to dive deep and read every story, they can get more expertise on an area or sector.”

She says the accuracy of the results is very high, and that “a tremendous amount of evaluation” was involved in developing the summaries, which included using its own newsroom as a testing ground to ensure journalists deemed them an accurate reflection of the full stories. Other factors that were evaluated include quality, timeliness, conciseness and completeness.

“There were different rounds of testing and evaluation before we were confident we could put it in front of users, because they’re making decisions based on the information,” she says. “Then, there has been a beta-testing period with some users, and we’ve spoken with them about how they engage with the summaries and what value they bring.”

One firm involved in that beta phase was proprietary trading firm Trillium Trading. In a statement, George Karas, senior executive trader at Trillium in New York, describes the summaries as providing “clear, concise insights that allow me to quickly grasp the essence of complex stories. As an event-driven proprietary trader, I find this to be an incredibly valuable tool for staying informed in a constantly evolving market.”

The summaries may appear differently depending on the type of story being summarized.

“Often, we think of news stories as straightforward articles related to a single event or development. But then there’s news analysis that connects the dots… and pulling information together. This is where summaries are especially useful because they can go in and elevate key data points,” Quinones says. “And then there are features, which have narrative structures, and summaries give you the birds-eye view of the salient points.”

Though the GenAI aspect is new—such as the vendor’s AI-Powered Earnings Call Summaries, launched almost a year ago—Bloomberg has been using AI to make clients’ work easier for long before the latest GenAI craze.

“My team has been at the top of this space as it relates to AI for 15 years,” Quinones says. “We’ve got a strong team that understands how to apply the technology. Our clients are trying to get the most information possible to make the best decisions.”