Banking and financial services executives have expectations of high AI payoff in 2027 but, in order to realize those ambitions, they’ll need to overcome a range of barriers created by stodgy operating practices including business siloes.
Those are top-level takeaways from new financial industry data published by IBM’s Institute for Business Value.
In analyzing the financial industry, IBM’s experts cited an “architectural mismatch between AI ambition and current operating models,” identifying the need to redesign the institutions themselves in order to accelerate greater levels of autonomy while maintaining trust, compliance, and security.
Banking on Success
Banking executives expect major impact from their work with agentic AI, including these expected gains in 2027:
- 61% improvement in ROI from AI investments
- 57% faster incident detection
- 49% reduction in transaction errors
- 24% improvement in compliance accuracy
These items represent an important cross section of critical vertical industry functions and processes. They also underscore agentic AI’s potential to realize reinvention — not just enhancement or incremental improvement.
IBM’s analysis said these figures highlight a potential “structural reinvention” in which institutions “shift from manual risk monitoring to more continuous autonomous surveillance; from batch-based finance to real-time financial intelligence…and from siloed analysis to interconnected enterprise
decision making.”
Financial services execs recognize there are significant structural barriers and limitations that stand in the way of reinvention. By their account:
- Only 26% of responding firms have global workflow owners with end-to-end accountability – a major organizational gap in view of agentic AI’s potential to automate workflows on an end-to-end basis
- 59% of respondents report they’re actively dismantling functional boundaries. Here’s one example based on their inputs: 30% of firms continue to operate through regional hubs with processes that are locally optimized rather than taking a broader perspective that could be national or international in scope.
- 79% say integrating data and technology across partners is a major challenge in their business – also a big barrier in light of the interconnected nature of financial transactions
IBM’s researchers draw an important conclusion about operating models’ role in creating AI impact: horizontal, interconnected workflows are “becoming a structural prerequisite for AI-enabled, real-time financial operations, not merely an efficiency preference.”
As they move forward in search of tangible business impact, high percentages of financial execs expect autonomous operations to play big roles in a range of processes and functions:
- 75% of firms are planning for autonomous operations in IT and data security/threat detection
- 65% in customer service and support
- 58% in financial planning workflows
- 55% in risk and compliance visualization models
As they move forward, financial execs will grapple with business challenges before delivering on their expected AI outcomes; some of those are industry-specific while others are more generally applicable across industries. They include:
- 75% cite data security and privacy as major challenges to AI-enabled operations
- 68% say they’re experiencing a talent shortage in AI expertise
- 60% identify regulatory scrutiny as a major concern
Needed: Tech Innovation + Reinvention
We have closely tracked the delivery of AI software designed with customer input, connectors, and pre-packaged AI skills from top AI vendors (see links below) focused on delivering tailored AI packages for financial services. Those solutions will better position financial institutions to deliver AI outcomes in core financial services functions.
Still, the IBM data makes clear that banks must overcome outdated operating models, siloes, and structural inefficiencies as well in order to make good on AI’s considerable potential. They’ve taken the first step in acknowledging the need for change, now the test will be whether they can effect required changes and position themselves to fully exploit the technology innovation that’s in the pipeline.
More Vertical Industry AI Insights:
- OpenAI Designs Software With Financial Customers, Embeds Third-Party Data
- Google Cloud’s 4-Point Plan for Industry Specific Success With Gemini Enterprise for Industries
