USA 
United Kingdom
BNY Mellon has revealed that dozens of its new AI-powered “digital employees” now operate with system logins and direct human line managers.
These tools are already performing tasks such as code writing and validating payment instructions across teams, according to Chief Information Officer Leigh-Ann Russell.
Similar to human workers, the AI-driven systems are assigned to specific teams, operate autonomously, and may soon gain access to tools such as email and Microsoft Teams for workplace communication, she said. “This is the next level,” Russell noted. “I’m sure in six months’ time it will become very, very prevalent.”
The move is part of a wider trend in financial services to embed artificial intelligence into core operations, with varying approaches to autonomy and system access.
Digital staff with autonomy and oversight
BNY Mellon’s AI Hub has created two digital employee personas, one focused on identifying and fixing code vulnerabilities, the other validating payment instructions. Each persona can exist in several instances, each embedded within a particular team. “That way no digital employee has broad access to information across the company,” Russell said.
Russell added that because these tools operate with logins and can access the same apps as humans, they are capable of working independently. For example, one AI system can identify a coding flaw, write the patch, and submit it to a manager for approval - all autonomously within the company's systems.
The company is exploring further use cases beyond code and payments and plans to extend communication capabilities for its AI agents, enabling them to contact managers directly through enterprise tools when escalation is required.
“We are working to build digital employees that go beyond coding and payment instruction validation into other areas,” the company said. It also confirmed it would continue to recruit “top human talent” alongside expanding its digital workforce.
AI staffing raises access questions
Some firms are taking a more cautious approach. JPMorgan Chase, for instance, is still defining access and management controls for similar AI tools. “It’s an open question exactly how much or how little access to give an agent,” said Chief Analytics Officer Derek Waldron, noting this may need to be decided on a case-by-case basis.
Waldron added that 230,000 JPMorgan Chase employees currently use a general AI chatbot via the company’s internal platform, with plans underway to develop increasingly tailored, autonomous versions aligned to different job roles.
He sees AI staff as a conceptual model to help business leaders understand how these tools function. “They are fundamentally different from human employees, of course, but also traditional software systems,” Waldron said.
AWS’s Managing Director for Financial Services, Scott Mullins, said the question of workforce integration is central: “How do we coordinate that work together? How do we manage those folks? How do we actually instruct those folks? What’s the new operating model? Those are the answers that we’re all working on right now.”
