US banking giant Citi has revved the Stylus Workspaces AI platform it has been rolling out to employees, touting that it is “now powered by agentic AI.”
The agentification of the platform, first announced last December, comes as analysts have begun taking a slightly more tempered view of the benefits AI can deliver to companies.
In a statement, Citi said it expects the the upgrade to help workers “tackle longer, more complex objectives and tasks with greater speed, efficiency and insight, supporting Citi’s efforts to become an even more technology-driven and productive firm.”
Tasks reportedly include researching a customer, constructing a profile of them from public and internal data, and converting it into a foreign language.
Stylus Workspaces integrates “directly with Citi systems, linking seamlessly to internal data and project management tools” as well as “commonly used enterprise platforms and other key resources.” And it can use “web searches and analysis capabilities.”
This means employees can “conduct in-depth research, extract insights from vast datasets and streamline multi-stage workflows into a single, automated process.”
CTO David Griffiths told the WSJ that Citi’s AI plaform uses a variety of models from Gemini to Claude, and has input limits into the system in a bid to prevent costs from spiraling.
The pilot involves 5,000 users and runs for up to six weeks, during which time the CTO will try to ascertain how impactful the tech is and calculate the return on investment.
“Does it mean that we need less people? I don’t know,” he said. “It certainly means that we would get a lot more done. And we’ll see how the workforce evolves with that massive boost of capacity that we’re getting here.”
In a statement, Griffiths said staff in the trial “can focus less on manual tasks and more on the big ideas that drive our business forward.”
There’s one manual task they might want to retain though – checking the output of agents. Back in June, The Register reported that Gartner expected more than 40 percent of agentic AI projects would be cancelled by the end of 2027.
The likely reasons for the expected pullback included rising costs, unclear business value, or insufficient risk controls. All of these should be anathema to bankers. At the same time, researchers highlighted a success rate for tasks carried out by agents of just 30 to 35 percent.
Earlier this month Gartner analysts highlighted the potential impact of AI in finance, and placed agentic AI right at the beginning of the famed Gartner hype cycle, describing it as an “innovation trigger.”
However, gen AI is now sliding down into the so-called “trough of disillusionment” though “Gartner experts… expect an impact on most finance functions within two years.”
Back in July Gartner said that Gen AI for procurement was also languishing in the trough of disillusionment.
“Many organizations are experiencing uneven [Return On Investment] or falling short of expectations, highlighting the need for a more measured and strategic approach,” the firm said.
Does that mean companies should be wary of adopting such overhyped technologies? Of course not.
Gartner declared: “Organizations that delay action on integrating GenAI into procurement processes risk falling behind as early adopters overcome these challenges and realize tangible benefits.” ®