The percentage will vary by process. A highly repetitive workflow may achieve much greater automation than a process requiring complex human judgment.
Challenges of Combining AI and Automation
Data Quality
AI models depend heavily on the quality of their data.
Incomplete or inconsistent banking data can produce unreliable results.
Automation can then make the problem worse if it automatically distributes incorrect information across multiple systems.
Security
Both AI systems and automated workflows may have access to sensitive financial information.
Banks need strong controls around:
- Authentication
- Access permissions
- Encryption
- Monitoring
- Audit logs
- Data protection
Model Accuracy
AI systems can make mistakes.
A model may incorrectly classify a transaction or customer request. For this reason, banks should establish appropriate monitoring and escalation procedures.
Legacy Banking Systems
Many financial institutions still operate complex legacy infrastructure.
Connecting modern AI systems to older applications can require APIs, integration platforms, or automation layers.
This is one reason banks may adopt automation incrementally instead of replacing their entire technology environment.
The Future of AI and Automation in Banks
The next phase of banking technology will likely involve increasingly connected AI and automation systems.
AI may become responsible for more sophisticated analysis, while automation coordinates actions across different banking applications.
The combination could support:
- More personalized customer experiences
- Faster fraud investigation
- Automated document processing
- Smarter compliance monitoring
- More efficient back-office operations
- Faster digital product development
The market is already moving in this direction. A 2024 Bank of England/FCA survey found that 55% of AI use cases reported by financial firms had some degree of automated decision-making, showing that AI adoption is increasingly connected with operational workflows rather than being limited to experimental tools.
Conclusion
AI and Automation in Banks work best when each technology performs the role it is designed for. AI can interpret information, recognize patterns, and support predictions, while automation can execute repetitive processes and move work between systems.
Together, they can improve banking operations across customer onboarding, fraud detection, customer service, compliance, and risk management.
The goal should not be to automate every banking decision. A more responsible approach is to automate predictable work, use AI where it adds analytical value, and keep people involved when decisions require judgment or accountability.
