How Major Banks Are Using AI to Replace Routine Work — and Why Regulators Are Raising Alarms

Its consumer virtual assistant, Erica, has handled billions of customer interactions since its launch and now processes tens of millions of requests each month within the bank’s mobile app . These tools answer questions, monitor spending patterns, and flag potential account issues automatically.

The bank has also rolled out Erica for Employees, an internal AI assistant used by nearly all of its more than 210,000 employees to handle routine tasks such as password resets, equipment requests, and internal support questions .

By shifting repetitive service work to AI systems, Bank of America aims to reduce operational workload for staff while improving response times for customers.

JPMorgan Chase: Hundreds of AI Use Cases

JPMorgan Chase is widely considered one of the most advanced banks in enterprise AI deployment. The firm operates hundreds of AI applications across its business lines, covering tasks such as coding support, compliance monitoring, data analysis, and research .

More than 200,000 employees have access to the bank’s internal large‑language‑model tools, which assist with document drafting, analytics, and workflow automation .

This large‑scale deployment reflects a broader strategy: AI absorbs routine knowledge‑work tasks so human staff can focus on higher‑value analysis and client interaction.

Goldman Sachs: Generative AI in the Back Office

Goldman Sachs is also rolling out generative‑AI assistants internally. The firm began deploying an AI assistant to thousands of employees in 2025, with plans for broader adoption across the company .

These tools help employees summarize documents, draft materials, and analyze financial data. Leadership has indicated that productivity gains from AI will likely slow hiring and gradually reduce some roles over time rather than triggering sudden layoffs .

Across Wall Street, the pattern is similar: fewer repetitive or junior roles, but rising demand for engineers, AI specialists, and model‑risk experts.

Why Regulators Are Becoming Concerned

Financial regulators broadly support innovation in artificial intelligence, but they are increasingly warning that governance and oversight may not be keeping pace with adoption.

1. Weak AI Governance

Australia’s prudential regulator, APRA, has warned that governance, risk management, and assurance practices inside financial institutions are not evolving quickly enough to match the scale and complexity of AI deployment .

Similarly, Australia’s securities regulator ASIC found potential governance gaps in how financial firms manage AI systems that affect customers, including how they assess risks such as bias or unfair outcomes .

2. Human Oversight and Explainability

Many AI systems used in finance—such as those for lending decisions, fraud detection, or compliance monitoring—can involve complex machine‑learning models that are difficult to interpret.

Regulators emphasize that firms must maintain human oversight and clear accountability when automated systems influence customer outcomes or regulatory decisions .

3. Cybersecurity and Operational Resilience

AI systems can also introduce new operational risks. Regulators warn that integrating AI tools into critical banking systems may expand cyber‑attack surfaces and increase dependency on external vendors or cloud providers .

Maintaining operational resilience—the ability of financial institutions to keep services running during disruptions—is therefore becoming a key regulatory priority.

4. Board‑Level Accountability

Regulators in several jurisdictions stress that AI risks remain the responsibility of bank boards and senior executives, even when systems are automated or outsourced.

In the UK, authorities such as the Financial Conduct Authority (FCA) have said existing regulatory frameworks—including consumer protection, model‑risk management, and operational resilience rules—already apply to AI deployments in financial services .

The Bigger Shift in Banking Work

Taken together, these developments signal a major transformation in how banks operate.

Artificial intelligence is becoming an operational layer across the entire industry, handling tasks that once required large numbers of analysts, operations staff, and support personnel. The shift does not necessarily eliminate human workers altogether, but it does change the mix of skills banks need.

Routine administrative and analytical work is increasingly automated, while demand grows for roles involving AI oversight, system design, cybersecurity, and client relationships.

The tension between innovation and risk is therefore becoming central to financial regulation. Banks see AI as a path to productivity and efficiency. Regulators see it as a powerful technology that must be carefully governed to protect consumers and maintain financial stability.