← Back to articles

Beyond Containment: Why Financial Services' AI Adoption Gap Remains

Rethinking the ROI of Banking Chatbots

For years, financial institutions have aggressively adopted chatbots with impressive penetration rates – reaching 64.2% in banking and insurance according to Parloa’s recent State of Agentic CX report. Yet resolution rates lag far behind at just 7.4%. Malte Kosub, CEO and co-founder of Parloa, argues this isn’t a technology problem but rather a consequence of how success is defined.

The Deflection Economy

Kosub points out that many deployments prioritized call deflection over genuine resolution – a strategic choice enabled by procurement practices that rewarded cost reduction above all else. Rule-based systems with no reasoning capability were often marketed as AI, masking their limitations.

This pattern reflects structural governance failures where IT departments focused on system uptime while customer experience functions measured satisfaction separately—with neither group primarily accountable for revenue outcomes. As Kosub puts it: “When the CFO is focused on headcount reduction and the IT department on system stability, genuine resolution falls through the gap.”

The Readiness Gap

The next frontier lies in agent-to-agent readiness, where personal AIs directly access enterprise systems—currently achieved by only 1% of institutions. Barriers include outdated authentication frameworks, transactional APIs, and inflexible legacy infrastructure.

Parloa’s clients like Allianz, Swiss Life, and Barmenia demonstrate that true resolution capability requires modernizing core systems alongside AI deployment. Those with updated policy and account infrastructures saw significantly higher success rates than those operating on legacy platforms.

A New Era of Customer Expectations

The skepticism around agentic AI is valid only if resolution rates fail to improve substantially within the next two years—currently sitting at 7.4% for AI interactions compared to 10% for human escalations.

Kosub believes customers now have higher expectations after experiencing AI in other domains. The tolerance for scripted responses has diminished, creating an opportunity for institutions that prioritize genuine customer outcomes over cost savings.

Written with the assistance of AI. Reviewed and edited by the AfricanCEO editorial team.

Source: thefintechtimes.com

Need AI, fintech, or digital transformation consulting? Talk to SoatDev →