Industry

AI-generated text

Agentic AI poised to reshape banking customer interactions

Agentic, or agent-based, AI systems are evolving from passive question-answering tools into autonomous agents that interpret requests, make decisions and execute tasks — a shift that could fundamentally change how banks and customers interact.

Agentic AI poised to reshape banking customer interactions

Artificial intelligence has moved beyond simple question-and-answer tasks: increasingly sophisticated agent-based (agentic) AI systems can interpret requests, make decisions and execute certain tasks autonomously. This capability has direct implications for the financial sector, where the relationship between banks and customers and the mechanics of customer service could be fundamentally altered.

What this means for everyday banking

In a conversation with Stadler Gellért, head of the TC2 Data & AI team, he explained that agentic systems can take over parts of banking services on behalf of customers: the system can parse requests, follow rules and automatically perform transactions or administrative actions in cases where processes are well defined. This could reduce handling times and improve efficiency, particularly for repetitive or rule-based tasks.

Security and regulatory considerations

The discussion stresses that strong security guarantees are required when deploying agentic solutions. Transparency of decision logic, strict management of access and permissions, and auditability — the ability to trace why and how the system made a particular decision — are all essential. Additionally, protecting personal data and ensuring compliance with financial regulations are critical.

Organizational changes needed

Exploiting the technology requires more than technical rollout: banks and related organizations must adapt internal processes and working methods. According to the conversation, this will involve retraining staff, creating new oversight roles and establishing operational controls that ensure agentic systems behave as intended and that human supervisors can effectively intervene in critical situations.

Why this matters now

Agentic AI is not only a technological upgrade: its adoption can bring efficiency gains, faster customer servicing and potentially improved user experience. However, the associated technical, legal and operational risks mean implementations must be planned and governed carefully.

Where to hear the full discussion

The conversation with Stadler Gellért is available as a podcast episode: it can be listened to on platforms such as Spotify and Apple Podcasts, and via the embedded player accompanying the article. The episode was sponsored by TC2.

Tags: data, telecom, banking, efficiency, customer service, client, ai, relationship, podcast, bank, agentic AI

(Cover image source: Portfolio)