Artificial intelligence (AI) holds great promise for enhancing accuracy and efficiency in the banking industry, yet there remains significant hesitation about fully embracing the technology, particularly in core business functions.

AI is becoming increasingly recognized as a vital tool for banks, especially when it comes to improving customer service, fraud detection, and managing internal operations. A recent survey by Dentons revealed that 74% of financial services professionals have already integrated AI for IT and cybersecurity, while 72% use it for customer service and support. Additionally, AI is making headway in areas like sales, marketing, research, and accounting, with 69% adoption across these sectors.

Despite these advancements, many banks still have reservations about implementing AI in key operational areas, particularly in their transaction departments. This hesitation has led to a lack of clear strategies for AI adoption. Dentons’ survey found that only 29% of financial services organizations had a formal AI roadmap in place by mid-2024.

The banking sector is at varying stages of AI adoption, with some banks still hesitant to integrate AI into their core functions. Common concerns include limitations in AI’s capabilities for certain tasks and an organizational culture that is too risk-averse to tolerate the potential errors that might arise from AI use.

Survey responses also highlighted some specific concerns. The most significant issue, mentioned by 57% of respondents, was the lack of human oversight, which could lead to errors and raise questions about accountability for mistakes made by AI. Other concerns included the risk of diminishing the human talent pipeline due to reliance on technology (52%) and gaps in skills within banks (49%). This indicates that many banks don’t yet feel equipped with the expertise to safely and effectively implement AI.

There is a general consensus that these concerns can be addressed through the development of internal frameworks, including strong governance structures that ensure transparency in AI usage and appropriate risk assessment and mitigation strategies. However, progress in this area has been uneven, partly because of the complexity of creating these frameworks in such a rapidly evolving field.

A common challenge among financial institutions is creating unified governance systems that cover different areas of the business, address the various AI tools being deployed, and manage diverse use cases with differing legal and risk concerns. One potential solution is to establish processes that can adapt over time as AI usage grows and regulations evolve.

A growing best practice in AI adoption is implementing triage processes to review and assess the risks of AI tools before they are fully adopted. Key factors in these assessments include the extent of external usage, the sensitivity of the data involved, reputational risks, and the potential consequences of AI-driven decisions.

In addition to triage and risk-rating, many large financial institutions are setting up “technology sandboxes” to test AI and other fintech tools. These environments allow for feedback and informed decision-making, helping organizations manage the complexities of AI adoption.

A cautious approach to AI can be beneficial, as it offers the chance to build an AI inventory as new tools are introduced. This inventory will likely become a legal requirement as AI regulations continue to develop.

However, there are also significant change management risks to consider, particularly the impact of AI on the workforce. Issues like job displacement and the need for effective retraining will be crucial for smooth AI integration.

Despite these concerns, the banking and financial services sector is increasingly aware of the need to embrace AI. According to Dentons’ survey, 78% of respondents believe that organizations that fail to adopt AI will struggle to remain viable in the future. This underscores the pressure on banks to bridge the gap between their AI ambitions and the actions they need to take to make them a reality.