
It's not science fiction! This bank has brought in dozens of AI employees to work alongside humans

Bank of NY Mellon has deployed dozens of AI "digital employees" with independent login credentials, working collaboratively with human employees; Goldman Sachs has introduced AI assistants to 10,000 employees, and JPMorgan Chase's 230,000 employees can use AI chatbots... The AI workforce in the financial industry has already reached the Next Level
When your next bank colleague could be a robot, science fiction is becoming reality.
Recently, Bank of NY Mellon announced the deployment of dozens of AI-driven "digital employees" to work alongside human employees, with these AI employees even having their own company login credentials. The Wall Street Journal reported that this marks an unprecedented new height in the application of AI in the banking industry.
The bank's Chief Information Officer Leigh-Ann Russell stated that, similar to human employees, these digital employees have managers to whom they report directly and can work autonomously in areas such as programming and payment instruction verification. She revealed that these AI employees will soon have independent email accounts and may even communicate with colleagues via Microsoft Teams.
According to reports, the AI Hub at Bank of NY Mellon developed two types of digital employee roles in just three months. The first type focuses on identifying and fixing code vulnerabilities; the second is responsible for verifying payment instructions. Each role can run multiple instances simultaneously (totaling dozens), but each instance is restricted to operate within specific teams to control access to company data. Russell pointed out that these digital employees have managers with direct reporting relationships, and at this stage, they mainly handle tasks autonomously. The bank emphasized that while expanding the scale of digital employees, its commitment to recruiting top human talent remains unchanged.
This technology is gaining increasing attention in the financial services industry. Russell stated:
This is the next level. I am confident that within six months, this (AI application model) technology will become very, very common.
Giants like Goldman Sachs are making moves
Goldman Sachs has launched an internal AI assistant for about 10,000 of its bankers, traders, and asset management managers.
The bank's Chief Information Officer Marco Argenti told CNBC that the assistant is currently used for basic tasks such as document proofreading and language polishing, designed to make users feel "as if they are talking to another Goldman Sachs employee." Argenti further envisions that the next phase will achieve "agent behavior," meaning the AI assistant can "complete" tasks like human employees, rather than just "express."
JPMorgan Chase Chief Analyst Derek Waldron views "digital employees" as a conceptual model to help business personnel understand AI tools. He believes that such digital employees are neither human nor traditional software systems and will require unique system connections and access management mechanisms.
Waldron envisions a future where every employee is equipped with an AI assistant, and every client receives AI concierge services. The company's 230,000 employees can already use a general AI chatbot through an internal platform, with the goal of developing more autonomous, intelligent, and deeply customized versions for different positions.
Challenges facing the industry
How to seamlessly integrate digital employees with human employee teams has become an urgent issue faced by financial institutions.
Scott Mullins, Director of Financial Services at AWS, stated that integrating digital employees with human employees is the primary issue facing the entire financial industry
How do we coordinate these tasks? How do we manage these employees? How do we actually guide these employees? What is the new operating model? These are all questions we are currently working hard to address.
Waldron also acknowledged that the exact amount of access to grant AI agents remains an open question that needs to be resolved on a case-by-case basis