U.S. fintech leaders are facing increasing pressure from their boards, competitors, and the news to implement AI. However, turning this hype into genuine return on investment (ROI) remains a complex and ongoing challenge.
Before requesting a significant AI budget, it is important to be aware of these three common artificial intelligence business case mistakes that frequently cause projects to fail before they begin.
1. Chasing AI Hype Over Proven ROI Opportunities
The most common mistake? Pursuing advanced, generative AI solutions while still struggling with manual, inefficient back-office operations. At CBTW Americas, we observe firms chasing media buzz rather than measurable profit.
Focus your AI implementation on proven, high-ROI areas:
- Fraud Detection & Risk: This is AI’s most mature application, representing over 33% of the market (Grand View Research). Machine learning slashes false positives and blocks fraud in real time.
- Customer Service Automation: Bank of America’s “Erica” has handled over 2.5 billion client interactions (Bank of America) – demonstrating scalable efficiency.
- Operational Efficiency (RPA): BNY Mellon improved processing time by 88% and achieved 100% accuracy for some tasks (Emerj). Start with internal processes for the fastest returns.
2. Overlooking Project-Killing Risks
Many business cases overemphasize upside, ignoring critical risks that can derail projects and budgets. Recognize and plan for these threats:
- Regulatory Complexity: Generative Artificial Intelligence enables new fraud risks. Deepfake-driven fraud could cost $40 billion in the U.S. by 2027 (Deloitte). Your business case must address governance, transparency, and compliance.
- The Talent Gap: In 2024, 100% of financial firms increased AI/ML investment – but talent shortages remain a major barrier (IIF/EY). Ensure your budget accounts for skilled staff, as they are as essential as the technology.
- Hidden Infrastructure Costs: AI requires powerful hardware and scalable hybrid-cloud infrastructure (Gartner). Insufficient hardware budgets often sabotage long-term scalability.
3. Going for a ‘Big Bang’ Launch Instead of Internal Wins
Everyone wants to unveil a revolutionary, customer-facing AI. However, leading U.S. financial institutions start by solving high-value internal problems with AI.
- JPMorgan Chase built an internal ChatGPT-like tool for wealth advisors.
- Morgan Stanley deployed a GPT-4-based assistant for internal data query needs.
These companies use GenAI to improve processes and prove ROI privately before going public (The UXDA). Meanwhile, firms investing in customer-facing chatbots often see limited results.
Actionable Guidance for Your AI Business Case
At CBTW Americas, we recommend you build the case around tangible ROI. Recent data reveals:
- 70% of firms attribute at least a 5% revenue boost to AI, and 60% report similar cost savings (Glassbox).
- Agentic AI shows 3.5 to 6 times higher ROI than legacy automation, with payback in under 14 months (Unique.AI).
The “next big thing” in finance – Autonomous and Embedded Finance – is coming (Devoteam, Chicago Partners). Don’t risk your roadmap by misallocating resources early. The U.S. financial sector is projected to spend $54 billion on Artificial Intelligence by 2028 (Statista), with up to $340 billion in annual value at stake (McKinsey).
Don’t get lost in the hype – invest where ROI is proven.
References
Bank of America, Chicago Partners, Deloitte, Devoteam, Emerj, Gartner, Glassbox, Grand View Research, IIF / EY, McKinsey, Statista, The UXDA , Unique
