AI Adoption and Strategy in Banking
📌 Bank leaders confirm that AI is a major focus, with roughly 70% of banks in the region engaging in some form of AI pilot projects.
🔬 A key challenge highlighted is that over 95% of Generative AI pilots have failed, often due to a lack of clearly defined outcomes.
💡 The consensus is that successful implementation requires defining clear outcomes first (e.g., improving customer experience, achieving operational efficiencies, or driving revenue uplift).

Consumer-Facing vs. Backend AI Applications
👩‍💻 Many current AI applications are backend-focused, but there is a strong push towards consumer-facing, multimodal interfaces using natural language (text, voice, images, gesture).
🤖 For banks like Right Bank, the goal is to bypass traditional UI constructs and replace them with natural language interaction, viewing this as a once-in-a-lifetime opportunity to reinvent human-computer interface.
📈 Traditional banks are focusing on using AI in the backend to improve turnaround times and support staff, adopting a "digital bank plus much more" approach that values human interaction.

Specific Use Cases and Quantifiable Results
🗂️ Collections: Hong Kong Bank saw AI-driven collection agents increase efficiency by 15% alongside an 86% reduction in associated costs by optimizing contact points (email vs. phone).
📈 Customer Engagement: Aon Bank saw active users engaging with their financial insights tool double (from 2-3 times/week to 6-7 times/week) after launching GenAI-enhanced personalized financial insights.
🛡️ Security & Operations: CIMB noted that GenAI chatbots contain 60% of support volumes, and AI-enhanced transaction/fraud monitoring reduced cycle time from 20-25 minutes down to 3-4 minutes.

Transforming the Workforce and Scaling AI
🧑‍💻 The nature of work is changing, with AI tools transforming roles from design and coding to compliance. One bank achieved a 2-week to 2-day reduction in answering 400-page RFPs using AI summarization.
🎓 Companies are promoting an "AI as a tool" mindset; in one large conglomerate, 90% of employees use AI for brainstorming, productivity, coding, and content creation.
🤝 Scaling AI requires decentralization: institutions are moving away from relying solely on central teams by launching internal labs and co-creation models to empower business units to build their own solutions.

Key Points & Insights
➡️ Define the 'Why': Do not implement AI just because it is trending; clearly identify the business outcome (CX, efficiency, revenue) before starting any project.
➡️ Start Small & Chunk Work: Avoid trying to think too big; focus on being very specific about the problem and breaking use cases down into achievable small chunks for rapid testing.
➡️ Embrace Multimodal Future: Expect customer interaction to become multimodal (text, voice, image), though the pace of adoption depends on wider industry acceptance and regulatory navigation.
➡️ Counter AI with AI: Recognize that threat actors are using GenAI; financial security strategy must evolve into an AI vs. AI battle for fraud detection and defense.

📸 Video summarized with SummaryTube.com on Nov 24, 2025, 07:19 UTC