It is early days for generative AI (GenAI) adoption at banks. And even earlier days for agentic AI adoption. Celent anticipates the current wave of GenAI adoption by banks will remain focused on advancing intelligent virtual assistants (IVAs) that respond to queries and requests from employees and customers. Celent believes that agent chaining (combining specialized agents to support a workflow or customer journey) will have profound impact in numerous areas of banking across lines of business.

But it’s not too early to monitor agentic AI trends and ideate use cases. Celent recommends that bankers develop their AI/GenAI strategy with the anticipation that agentic AI-supported workflows will become a viable option in the next five years. They should ideate internally and engage with third-party providers of agentic AI solutions, monitoring their progress and that of early adopter banks.

To assist our clients to grasp the meaning and implications of agentic AI in banking, Celent has compiled a flash report. It goes from the 10,000 foot level (definition) to ground level (examples and tech providers).

Figure: The Agentic AI Revolution: from Say to Do and Think

Companies mentioned in this report include: AWS, Google, IBM, Oracle, and Microsoft; Anthropic, Cohere, Meta, and OpenAI; crewai, Hugging Face, and LangChain; Glean, Inflection AI, and Relevance AI; Salesforce, ServiceNow, Nvidia, and UiPath; Goldman Sachs.