American Express is deploying AI to overhaul its customer service model, replacing traditional surveys with real-time sentiment analysis across every interaction.
Speaking at Money 20/20 USA on Monday (27 October), Mohammed Badi, President of Global Servicing, revealed that Amex (pictured, right) now uses machine learning to evaluate tone, intent, and satisfaction during live customer engagements.
“We no longer rely on outdated end-of-interaction surveys,” Badi said. “AI allows us to measure every single customer interaction in real time, analyse sentiment, and identify opportunities to improve the overall experience.”
The system enables Amex to pinpoint friction points, flag service gaps, and proactively coach agents, while building a dynamic feedback loop that evolves with customer behaviour.
Badi emphasised that this shift gives the company “a holistic view we simply couldn’t have before,” allowing for faster resolution and deeper personalisation.
This AI-driven approach reflects a broader trend in financial services. According to a recent report by McKinsey, banks and payment firms that integrate real-time analytics into customer service see up to 30% improvements in satisfaction and retention.
Bloomberg notes that AI adoption in financial servicing is accelerating, with firms like JPMorgan and Capital One investing heavily in conversational AI and predictive support tools.
American Express’s strategy positions it at the forefront of this transformation, where data isn’t just collected, but actively used to shape the customer experience in the moment.
Eric Sagger, COO at Plaid was also on stage for the fireside chat titled “Beyond Bots: Creating Meaningful Customer Experiences with AI.”
While Badi opened the discussion by reflecting on American Express’s 175-year legacy, Sagger emphasised Plaid’s commitment to transparency and control in data use, noting that the company’s systems are “fast by design” to support millions of users globally.
Plaid is a fintech company that connects consumer bank accounts to financial apps.
“We’re explicit about how we use data,” Sagger said.
“By clearly asking customers for consent to share data, we maintain both control and visibility. Transparent data usage is essential for building trust – and for enabling agentic AI that genuinely improves customer experiences without compromising safety or innovation.”
Badi echoed the sentiment, highlighting that trust and brand are synonymous at American Express.
Across the session, a clear message emerged: AI, when used responsibly and transparently, can deepen customer trust and deliver richer, more human experiences.