Rudy Lai

AI @ American Express

Payments and financial services
Industry
Last updated
July 3, 2025 at 10:44 AM

Summary

  • American Express has steadily increased its AI adoption since 2021, evolving from traditional machine learning models for credit checks and fraud detection to cautiously experimenting with generative AI and large language models (LLMs) by 2023-2025.
  • Significant initiatives include AI-powered customer service virtual agents supporting multiple languages, advanced fraud detection led by experts like Dmitry Efimov, and AI solutions improving travel management and IT operations in American Express Global Business Travel (Amex GBT), demonstrating operational efficiency and enhanced customer experience.
  • By 2025, American Express has operationalized over 70 AI use cases, emphasizing predictive analytics, agentic AI-assisted self-service solutions, and innovation governance programs, with leadership figures such as Luke Gebb and Ravi Radhakrishnan driving AI strategy and innovation.

VIBE METER

More AI announcements = more VIBE
Q1 23Q2 23Q3 23Q4 23Q1 24Q2 24Q3 24Q4 24Q1 25Q2 25Q3 25Q4 25🔥🔥🔥🔥

5 AI Use Cases at American Express

Customer Service
2025
Customer Facing
Traditional
Generative
Agentic
Outcome
Amex GBT uses large language models and natural language processing in multilingual virtual agents to handle customer inquiries autonomously, increasing self-service rates and improving customer experience. [1][2]
IT Operations
2025
Traditional
Generative
Agentic
Outcome
Costs
An AI-powered IT chatbot autonomously resolves common technical issues, markedly lowering the volume of escalated IT support tickets and enhancing internal operational efficiency. [1]
Travel Support
2024
Customer Facing
Traditional
Generative
Agentic
Outcome
AI tools assist Amex GBT travel agents with complex travel policy understanding and automate mundane tasks, enabling agents to focus on higher-value customer service activities. [1][2]
Fraud Detection
2023
Traditional
Generative
Agentic
Outcome
Risk
American Express employs advanced AI models and decades of experience to identify and prevent credit card fraud, significantly reducing risk and losses by analyzing transaction patterns. [1][2]
Credit Approval
2021
Traditional
Generative
Agentic
Outcome
Costs
Using machine learning, American Express speeds up the commercial card approval process by quickly assessing creditworthiness and automating decisions that traditionally required manual review. [1]

Timeline

2025 Q4: no updates

2025 Q3

1 updates

American Express continued solidifying AI strategy centered on leveraging proprietary data and closed-loop network advantages to dominate financial services AI usage.

2025 Q2

1 updates

Amex GBT integrated LLM-powered multilingual virtual agents and natural language query analytics tools, enhancing customer self-service and operational efficiency; leadership emphasized AI adoption maturity with over 70 live use cases.

2025 Q1

1 updates

Amex GBT invested in AI fluency programs for employees to foster innovation and enhance AI adoption culture and capabilities.

2024 Q4

3 updates

American Express Ventures championed generative AI experimentation focused on productivity, protection, and growth; Amex GBT revealed first AI tools aiding travel agent onboarding and travel policy comprehension.

2024 Q3

1 updates

Amex GBT utilized AI to automate trip request responses via email, aiming to augment human agents by freeing them for high-value tasks in business travel management.

2024 Q2

2 updates

Amex disclosed positive engineer feedback on generative AI 'copilots' assisting coding; Amex GBT aggressively piloted AI adoption across various departments under executive oversight.

2024 Q1

2 updates

Amex Global Business Travel launched an initiative to drive AI innovation and governance, aiming to unlock efficiencies and enrich client experiences, while joining Stanford HAI's corporate affiliate program.

2023 Q4: no updates

2023 Q3

2 updates

American Express continued AI integration emphasizing automation and decision support, illustrating the balance between innovative tech and established practices.

2023 Q2

4 updates

American Express cautiously experimented with generative AI technologies such as ChatGPT, focusing on sentiment analysis and enhancing fintech capabilities without plans to develop proprietary large language models.

2023 Q1

2 updates

Advanced AI models demonstrated progress in credit card fraud detection led by Dmitry Efimov; marketing operations leveraged AI for enhanced decision-making and campaign effectiveness.

2022 Q4: no updates

2022 Q3: no updates

2022 Q2

1 updates

American Express educates on AI technology benefits including chatbots and supply chain management algorithms targeting business improvements.

2022 Q1: no updates

2021 Q4

1 updates

Further exploration of AI applications in business and industry, highlighting unique use cases shaping American Express's technology approach.

2021 Q3: no updates

2021 Q2

2 updates

American Express applied AI to enhance secure transactions and improved credit approval processes with extensive machine learning enabling faster commercial card issuance.