AI for Finance Summit by Artefact - September 17th, 2024 - Paris
Key learnings from the panel discussion with Ghislain Lefebvre, Senior Strategic Account Executive, South Europe at Treasure Data, Guillaume Rincé, CTO at Maif, and Sophie Huss, Chief Marketing Officer at Artefact.
MAIF’s AI approach and responsibility
MAIF has been using AI for nearly a decade, focusing on boosting company performance and supporting employees. The company takes a responsible approach to AI, aiming to protect client data and minimize environmental impact. As a socially responsible insurer, MAIF aligns AI strategies with its mission to protect the planet, balancing efficiency and sustainability.
AI industrialization and generative AI governance
Traditional AI is already industrialized at MAIF, powering tools like email routing and voice systems for customer service. Generative AI has brought new opportunities, prompting extensive internal experimentation throughout 2023. After identifying business expectations, leadership prioritized key AI applications to boost company and client relations, aiming to scale up select use cases effectively.
MAIF’s key AI use cases
MAIF leverages AI in practical ways: summarizing client conversations for better advisor follow-ups, aiding legal teams in research and analysis, and making internal knowledge bases more accessible through conversational AI. These efforts improve efficiency and support client-facing and internal teams.
Treasure Data’s contribution to consumer data use
Treasure Data supports large enterprises by consolidating consumer data across departments, enhancing personalized experiences. Working in finance, including partnerships with AXA and SMBC in Japan, they accelerate campaign creation and expand activation channels. This has led to improved engagement rates and reduced campaign design times, all supported by AI-based solutions.
Data sovereignty and AI at MAIF
Guillaume Rincé emphasizes the need for data sovereignty, ensuring that client data stays in Europe and is not used for non-European AI model training. The importance of culturally relevant AI models is noted, as European, American, and Chinese consumer behavior differ greatly. To ensure sovereignty, MAIF is exploring open-source models and datasets based on European content.
Security and compliance at Treasure Data
As an American provider, Treasure Data is highly focused on data security, with hosting on AWS European data centers and compliance with European standards. Their solutions are designed to be secure, auditable, and suitable for large corporations, particularly those in regulated industries like finance.
MAIF’s governance for responsible AI
MAIF has expanded its cloud governance to cover AI governance, ensuring transparent, collective decision-making. Policies and decision trees are built to clarify technology choices and support regulatory compliance, reflecting the company’s commitment to responsible AI.
Balancing performance and sustainability
Guillaume Rincé stresses the need to balance technological benefits with environmental responsibility. European companies must make thoughtful decisions to support the growth of European AI while considering the environmental footprint. By selecting AI models with the right balance of performance and sustainability, MAIF aligns its technology choices with its mission.
A responsible AI strategy
MAIF’s AI governance aligns with its social values and mission, making responsible choices even when they may limit performance. This approach supports a broader movement for citizen-focused, responsible AI in finance.