AI for Finance Summit by Artefact - September 17th, 2024 - Paris
Key learnings from the panel discussion with Jean-Philippe Desbiolles, Managing Director, IBM Financial Services at IBM, Damien Ernst, Deputy CIO and CTO at Crédit Mutuel Alliance Fédérale, and Emma Sezen, Head of AI for Finance at Artefact.
Collaboration between Crédit Mutuel and IBM
Since 2015, Crédit Mutuel Alliance Fédérale and IBM have collaborated on AI, starting with analyzing advisor emails and providing virtual assistance. A cognitive factory was established to support the implementation of AI use cases, involving IT, business experts, and IBM support. In 2018 they integrated AI across various activities, and began exploring generative AI in 2023.
AI strategy and scaling
AI strategies require sticking to a long-term vision despite challenges, rather than “failing fast.” True AI at scale involves creating an operational model, like an AI factory, and tightly integrating AI with data to achieve tangible results. This goes beyond technology to require organizational change and the right platform to ensure successful AI adoption.
Trust and ethics in AI
In finance, trust is vital. Crédit Mutuel Alliance Fédérale emphasizes ethics and customer protection, centralizing AI models and data for auditability. They adopted IBM’s WatsonX for monitoring model transparency, compliance, and ethical principles. IBM adds that trust drives adoption and ROI. Establishing a platform to centralize AI models, ensuring data integrity, and developing explicit ethical guidelines are crucial steps for responsible AI.
AI’s impact on the bank’s operations
AI is embedded across many business functions, bringing efficiency and freeing up time for value-added customer interactions. Tools enhanced by AI handle over 6 million customer emails monthly, analyzing attachments, supporting advisors with 600,000 monthly queries. Customers ask over 1.4 million questions through AI-powered virtual assistants, and an additional 3 million calls are managed by AI-enhanced voice systems. These improvements led to freeing 2.4 million hours for employees.
Challenges and future steps
While AI has improved operations, challenges remain. The bank focus is on integrating generative AI effectively, while ensuring technology remains a tool for enhancing human work, not replacing it. Cross-channel AI capabilities are another goal, ensuring consistent experiences across all customer touchpoints. Ethics, data privacy, and control over AI operations are priorities, as all data and processes run in Crédit Mutuel Alliance Fédérale ’s data centers, with private cloud capabilities enriched by GPUs for efficient AI processing.
AI and human collaboration
The primary challenge is the balance between humans and AI. Decisions made by AI alone are not always superior; the context of the use case determines whether AI, humans, or their combination provides the best outcome. As generative AI evolves, organizations must prepare to address questions around decision-making accountability, especially in sensitive fields like healthcare and finance.
The next frontier
Quantum computing: The banking industry’s next step is combining AI, traditional IT, and quantum computing. Crédit Mutuel Alliance Fédérale is exploring quantum computing for its potential to enhance operational efficiency, having established a Quantum Factory and Academy to develop expertise in this field. Future IT will converge traditional computing (bits), quantum technology (qubits), and AI (neurons).