JPMorgan Chase, the world’s largest bank by market cap, began rolling out to employees its own internal chatbot, known as LLM Suite, per a report in the Financial Times. 

They were encouraged to use it for “writing, generating ideas, solving problems using Excel, summarizing documents,” among other things, according to an email sent by the bank. 

At JPMorgan, the chatbot could augment the work being done with “a hybrid of human and AI analysts, very similar to how the intelligence community works,” Igor Jablokov, founder and CEO of AI startup Pryon told Fortune. 

JPMorgan employees got access to the chatbot in piecemeal fashion. An email sent to employees in JPMorgan’s asset and wealth management division announced they would be the latest to receive access to LLM Suite. So far, it has been rolled out to roughly 50,000 of JPMorgan’s 240,000 global employees, according to the Financial Times. 

“Think of LLM Suite as a research analyst that can offer information, solutions, and advice on a topic,” noted the email, portions of which were viewed by Fortune. The memo referred to LLM Suite as a “the firm’s ChatGPT-like product.”

On Wall Street, AI has been met with both anticipation and trepidation. The industry could be poised to reap huge profits from AI’s potential, while at the same time it could be among the hardest hit by job losses as a result of automation. A study from consulting firm Accenture found bankers could have three-quarters of their daily tasks replaced by AI. And Citigroup, a Wall Street giant in its own right, forecasts the banking sector could lose more jobs to AI automation than any other. 

That said, companies aren’t stopping its integration into their day-to-day operations. As AI becomes increasingly accessible, companies are looking for ways to tailor it to their own specific needs. Financial firms in particular have been excited about generative AI because of its ability to sift through huge amounts of data and make connections that may have otherwise gone unnoticed by humans, according to Matt Lucas, field chief technology officer at startup Stardog, which recently developed its own wealth management AI tool.

JPMorgan’s chatbot was “not surprising,” he said in a phone interview. “There’s definitely an appetite for introducing this type of capability inside of firms.”

LLM Suite is not JPMorgan’s only AI chatbot. The bank also has two other tools known as Connect Coach and SpectrumGPT that are specific to business tasks, rather than a general purpose tool like LLM Suite. 

JPMorgan declined to comment. 

“Anytime this type of technology gets introduced, there’s always the question of, ‘Is this going to replace the line worker or replace the advisor?’” said Lucas, who was executive director of technology at Morgan Stanley before Stardog. “I see it completely differently. It’s going to make them more effective and more powerful in their role.”

Lucas’s comments belong to a popular school of thought in the current AI boom. There have been studies from universities like MIT and Stanford pointing to the fact that employees who use AI are more productive. 

But that’s not to say AI won’t eliminate jobs at all. A survey of business executives found that 41% of them expected workforces would shrink because of AI. AI pioneer Mustafa Suleyman, cofounder of DeepMind, said that he saw the technology as “fundamentally labor replacing.” 

A widely cited Goldman Sachs analysis from May 2023 estimated 300 million jobs globally could be exposed to AI. Still, Goldman didn’t say all those jobs would be eliminated altogether, instead noting most would simply be reshaped, with only a portion of their work being done by AI. 

“Most jobs and industries are only partially exposed to automation and are thus more likely to be complemented rather than substituted by AI,” Goldman Sachs wrote. 

With AI, bank analysts could be able to access and synthesize far greater amounts of data, according to Alex Ratner, CEO of Snorkel, a startup that specializes in labeling enterprise data for use in large language models. Banks have long used various forms of machine learning to both sort data and make investment decisions. However, with generative AI those capabilities will become more accessible and, crucially for rank-and-file employees, applicable to mundane everyday tasks. Indeed, a PowerPoint from JPMorgan’s Investor Day in May referenced using AI tools to “remove ‘no joy’ work.” 

As workers’ jobs continue to change, a common refrain is that workers are more likely to be replaced by people adept at using AI, rather than an AI itself. Jablokov expects that jobs will change with AI, in the same way they did when computers became commonplace a few decades ago. 

“Remember that at the dawn of the Information Age the only middle managers that lost their jobs were the ones that couldn’t figure out how to use computers in their workflows,” he said. “It’ll be a similar form of top grading for these roles.”