AI is moving towards its agentic phase where AI usage across companies will become pervasive. But, there is still a long way to go before this matures and a lot of businesses need to sort out their data first. They need to look at their own standard canonical data model for their business, the cleanliness and accuracy of their data, and how they manage data on premise and/or in the public cloud before embarking on building trust in simpler AI models where ROI and business case are the biggest challenge to get right to ensure success
With the prevalence of smaller, more IP optimised, models becoming pervasive IBM are extraordinarily well placed to deliver real value to clients quickly. It’s IBM’s end-to-end strength across their business that positions them so well here; from embedded AI infrastructure for inferencing to contextual acceleration in hardware alongside, what will become a critical differentiator AI governance.
Large banks mission critical platforms are predominantly based on IBM Z and the release of the Telum chip in the IBM Z enabled significantly enhanced inferencing which has driven benefits for banks and software vendors.
Vendors taking advantage of this AI enabled architecture are DXC who are using the platform to reduce fraud; Worldline financial services who are using IBM Z to deliver enhanced merchant pre-payment; FIS who are using the platform to deliver dynamic overdrafts; eXate are using it to deliver improved regulatory governance; Ovation CXM have leveraged the platform to deliver improved merchant onboarding and activation; and Coforge are utilising the architecture to deliver improved security and efficiency for their banking clients. These use cases are outlined in more detail in the whitepaper, a link to which can be found at the bottom of this blog.
There is a direction of travel towards growth of on premise & sovereign AI. Security is fundamental to well governed AI. There’s also a need for greater energy efficiency for models. Combine these elements and it’s obvious that IBM Z meets these market requirements well and can deliver on all these areas with the most secure platform, the ability to drive the best performance in real time, where the operational data resides, at the most power efficient standpoint, whilst securing that data.
With the advent of the IBM Telum II Processor and IBM Spyre Accelerator on the next generation of IBM Z, clients can now combine unstructured data from LLMs on platform or from other hybrid cloud sources to deliver meaningful AI capabilities for mission critical applications that need to leverage value from their operations core data alongside unstructured data from other LLM models. IBM Z remains essential to financial services, with over 70% of global transactions running on IBM Z [1]. Its reliability and security make it the foundation of mission critical banking infrastructure. Financial institutions rely on IBM Z for transaction processing, but moving data off IBM Z for AI-driven insights presents challenges:
- Data summarisation – Valuable granular insights are lost in the ETL and/or ELT process
- Reduced data fidelity – Limiting real-time AI use cases and decision-making
Since banks require real-time anomaly detection rather than just aggregate data, running AI models directly on IBM Z provides a competitive advantage, especially for large banks.
AI use cases in banking
AI use cases in banking are far-reaching, with software and ecosystem vendors developing products with AI capabilities baked in. Some of the most impactful use cases include AI-powered digital agents, such as chatbots, which provide real-time customer support and streamline interactions. Additionally, AI can develop personalised marketing and communications by analysing customer data to deliver tailored recommendations and engagement strategies. Regarding security, AI-driven risk management and fraud detection systems help financial institutions identify and mitigate threats in real-time.
Supporting these advancements, IBM Z enables banks to integrate these AI solutions into their operations, ensuring high-performance, security, and scalability for AI-driven automation.
As AI advances IBM Z is Guiding AI transformation towards agentic and multi-model execution. IBM is uniquely positioned to continue supporting banks, financial institutions, hyperscalers, and fintechs in navigating this transition, ensuring AI-driven innovation while maintaining governance and security. AI on IBM Z and LinuxONE Discovery Workshops represent IBM’s investment in partnering with clients, demonstrating the path to get started with AI on IBM Z and LinuxONE today.
Author:
John Smith of Quantum Six Consulting
https://www.linkedin.com/in/johnsmith4
John heads up Ecosystem Advisory, Alliances and Marketing at Quantum Six
John created the Quantum Six Ecosystem Banking BlueprintTM (EBB) which is a market leading vendor selection asset that banks, Fintechs and Hyperscalers use to select the ideal banking ecosystem, accelerating time to contract, time to procure or time to select.
John’s team focus on helping vendors on how to build the business case, requirements, selection criteria, vendors, negotiation, procurement and 3rd party due diligence required to deliver best value to any financial organisation, Fintech or Hyperscaler for any business need.
The EBB has a significant amount of data points on AI based vendors and Quantum Six regularly advise clients on how to pragmatically and practically build the AI business case and drive successful adoption, and execution, of AI in the most cost effective manner.
Download the full report here
https://quantumsix.com/ai-powered-banking-transformation/
[1] Based on a report commissioned by IBM with Celent. Operationalizing Fraud Prevention on IBM z16, April 2022, page 3