AI Agents Enter Banking Roles at Bank of America

Summary

Bank of America is deploying an internal AI-powered advisory platform to a subset of its financial advisors — approximately 1,000 in total, according to Banking Dive. The platform is built on Salesforce's Agentforce, a solution that enables the creation of AI agents to handle tasks, and is designed to help advisors respond to client queries, prepare recommendations, and manage daily workflows.

This represents one of the clearest examples to date of AI being used in core banking functions, rather than in limited pilots or back-office automation. The move reflects a broader shift across the industry: AI systems are transitioning from basic assistants to tools capable of supporting real-time decision-making.

Bank of America already had an extensive track record in AI adoption, according to the same source: its virtual assistant Erica handles work equivalent to approximately 11,000 employees, and the bank's 18,000 software developers use AI coding tools that have improved productivity by around 20%.

In practice

What distinguishes this deployment from earlier generations of AI in banking is its positioning: rather than handling simple queries or automating routine tasks, the new system is embedded within the advisory process itself. This means the AI can analyse client data, suggest next steps, and support the preparation of recommendations — all within the advisor's workflow.

Financial advisors sit at the centre of the bank's relationship with clients, particularly in wealth management. Introducing AI into that role implies a level of trust in the technology that, until recently, was absent from banking deployments.

The system is not replacing advisors. The design is hybrid: human judgement remains central, especially for complex financial decisions or high-value clients. What changes is that AI now handles a greater share of analytical and preparatory work, potentially freeing advisors to focus more on client relationships.

Other large banks are following similar paths. Banking Dive reports that JPMorgan, Wells Fargo, and Goldman Sachs are also testing AI tools to improve productivity and support client-facing roles, though each institution is taking its own approach and not all are focused specifically on advisor-oriented AI agent systems.

Who benefits / who loses

Banks and their shareholders are the most immediate beneficiaries: the technology allows institutions to increase advisory capacity and productivity without expanding headcount at the same rate. According to deployment data cited in the article, there are gains in how quickly advisors can access information and prepare for client meetings, though results vary.

For financial advisors, the impact is more ambiguous. On one hand, AI can reduce repetitive research and analytical work. On the other, if systems take on more of the analytical load over time, the skills profile of the role may shift — placing greater value on relationship skills and less on technical analysis.

The risks are real. Errors in data or model outputs could affect recommendations. Overreliance on automated systems may reduce critical review by human staff. Financial regulation adds another layer of complexity: institutions must ensure AI-driven recommendations meet compliance standards and can be explained to regulators if challenged — which tends to limit the degree of autonomy provided to AI systems, particularly in areas such as lending or investment advice.

Wells Fargo analyst Mike Mayo, cited by Banking Dive, cautioned that recent developments have yet to produce major new products, describing the current phase as "a little boring from a product standpoint" — suggesting that, despite intense activity, the visible impact on customers is still materialising.

Some estimates cited in the article suggest that up to one-third of banking jobs, or parts of those roles, could eventually be handled by AI, though timelines remain unclear.

Why it matters

• Bank of America is among the first major banks to integrate AI agents into direct client-facing roles — not just back-office tasks or limited pilots — making this case a meaningful indicator for the rest of the sector.
• The choice of Salesforce's Agentforce as the platform's foundation confirms that major financial institutions are turning to established AI agent infrastructure vendors rather than building everything in-house.
• The hybrid human-AI model emerging here — where human oversight remains central — may become the operational standard for retail banking and wealth management in the coming years.
• Regulatory and compliance challenges inherent in using AI for financial recommendations will determine how far these deployments can advance in terms of autonomy and scale.