Meeting Summary
The IBM Financial Services Cloud Council Americas convened on September 28, 2023 to continue discussions and collaboration on the ways to de-risk cloud adoption and to accelerate digital transformation for the financial services industry. The meeting focused on three areas: Regulatory Update; FinOps: Cracking the Challenge of Delivering Business Value Across Enterprise IT, and Generative AI: Taking Advantage and Taking Control.
Mark Carioto, VP IBM Cloud, Americas opened the session by welcoming the group and the new members and provided an update on the Council’s activities and progress in 2023. The Council held in-person meetings in Madrid and New York City where Council members, policy leaders, and regulatory authorities engaged in dialogue around advancements in cloud, quantum computing, and how to drive innovation with security challenges. During these sessions, the European Central Bank, Bank of Spain, the European Commission, OSFI, FDIC, US Treasury, and OCC offered insights around enhanced expectations on third-party risk management and cloud.
The Council’s 2023 working group initiatives extended discussions from the Council meetings into deep dives around testing operational resilience and concentration risk. These initiatives resulted in tangible assets including a recently published whitepaper, Testing Operational Resilience, and a forthcoming whitepaper on Concentration Risk. Additionally, in October, the Council will convene a working group to focus on AI and the controls required to support the necessary guardrails.
David Kliemann, Cloud Risk and Controls Leader IBM Cloud for Financial Services updated the Council on discussions with regulatory authorities in the United States, Canada, Australia, and the United Kingdom. IBM Cloud is seeing a consistent approach with a focus on third-and fourth-party risk, especially SaaS and cloud providers; an expectation of strong governance for cloud migrations, ongoing monitoring for critical service providers; and direct supervision of Cloud Service Providers by competent authorities.
The U.S. Treasury has launched multiple cloud focused workstreams since the release of the U.S. Treasury Cloud Report in early 2023. IBM Cloud will be serving as an advisor and deputy co-chair of the Financial Services Sector Coordinating Committee’s (FSSCC) Cloud Executive Steering Committee Group. The FSSCC is comprised of the OCC, CFPB, and the U.S. Treasury, along with financial institutions. Perspectives from the Council will be brought forward to the committee and progress and updates from the FSSCC will be shared at future Council sessions.
FinOps Cracking the Challenge of Delivering Business Value Across Enterprise IT
Sarah Fiihr, Director, IBM Cloud GTM Product Management & Training opened the session by hearing from the Council members about the issues they are facing and the approaches they are taking around FinOps. Sarah started the dialogue with the challenges that most clients are facing today:
- Managing heterogeneous workloads
- Combination of cloud and on-prem staff
- Multiple providers across multiple CSPs
- Multiple SaaS and on-prem vendors
Sarah shared findings from Apptio customer survey data which can be found here as well as evidence from IBM clients. She highlighted a unique approach of partitioning spend by application, with spend management effectively becoming part of product management through the P&L owner. This resulted in clear accountability around cloud costs, as well as personal accountability in the team. The Council engaged further in the findings and discussed the approaches they are taking, what they are experiencing in their organizations, and the success and challenges they are encountering.
Generative AI: Taking Advantage and Taking Control
Jay Jubran, Director of IBM Cloud Product Management and Judith Pinto, Managing Director Promontory Financial Group opened this segment with two different points of view. Jay’s perspective focused on the innovation of Generative AI; what’s coming, what's changing, what IBM sees in the market and Judith covered the topic from a governance and managing compliance perspective.
Jay introduced and provided an overview of IBM watsonx which consists of three product sets; he also shared use cases within the financial services industry:
- watsonx.ai: training, validating, tuning, and deploying AI models.
- watsonx.data: scaling AI workloads, for all of one’s data, anywhere.
- watsonx. governance: enabling responsible, transparent, and explainable data and AI workflows.
Judith shared the regulatory perspective on Generative AI and how regulatory bodies worldwide are developing guidelines to enable safety, transparency, traceability, and non-discrimination. The regulatory landscape is evolving with a focus on the management of AI-related risk including potential applicability relevance of existing supervisory guidance or expectations. Model risk management, third-party risk management, new and modified products and services, risk management (appetite, monitoring, and testing), and compliance risk management are areas that regulatory bodies are scrutinizing.
Current and Completed Activities
Dave Kliemann reminded the Council to join the newly launched Financial Services Cloud Council & Forum (FSCC&F), here. In the online community you can find blogs, discussion documents, details of upcoming events, and whitepapers including the recently released Testing Operational Resilience whitepaper here. In addition, Dave reminded the Council of the upcoming Generative AI Controls Working Group on October 23/25 followed by the Cross Council Event on November 2.
Council Discussion - Summary of Key Themes:
Challenges of FinOps
“The complexity of everything working together, getting everything to work consistently and perform reliably in the cloud is the biggest challenge right now.”
“We are struggling because we're trying right now to apply the models that we used to manage and understand the cost of what is used in our data centers and trying to apply some of those same concepts to the cloud.”
“It's not necessarily different tools different outputs, it's different environments requiring different tools that don't necessarily give the same outputs when you're trying to provide it to regulators or auditors.”
- FinOps enables organizations to validate ownership of cloud spend, optimize costs, and improve ongoing savings.
- 45% of organizations have no formal way of measuring IT value. Some organizations lack alignment between the business and IT teams. The trends show organizations wanting to self-fund digital transformation, but not always understanding what tools they need to leverage to be cost effective.
- To solve for some of these challenges the answer is really in the data. The data is the life blood of FinOps. FinOps allows organizations to understand the data and make better decisions.
The Rapid Pace & Challenges of Gen AI
“One of the biggest problems for us is actually explaining things to regulators - it’s not just a matter of transparency, they want transparency … they want us to explain what it is in our algorithms, how we do the adaptation, how do we actually even run basic functions.”
“I think it's still too risky to put generative AI in front of your customers because again the answers could be wrong. How do you filter it, how do you make sure it's providing the right answers and how do you ensure that the customer sentiment will be ok. I think most of the applications we're using today is increasing productivity in the back end with AI such as KYC.”
- Banking and other industries are trying to understand the cost components of generative AI and the value AI will deliver. It was suggested, to take a public level model approach and fine tune it with your own data and that then becomes your own model. This model enables companies to own the assets that create the value.
- There will be a significant economic impact in terms of productivity with generative AI, estimated in the region of $16 trillion which is bigger than the EU GDP. It is going to impact every industry from banking to healthcare to retail and it's going to impact different workflows as well. It will also affect how institutions operate and how they interact with employees, clients, and suppliers.
- The focus areas for establishing AI governance are going to be critical moving forward: establish a program blueprint, enable and operate within a framework, and develop a ongoing governance program.
- Leading organizations have created a Center of Excellence approach with regular reviews, sharing of knowledge and best practices to support teams, and in-house education to enhance skills. These are all important elements for success.
Next Steps and Actions
- Generative AI Controls - join the Working Group to continue the discussion, scheduled on 23/25 October.
- Cross Council Event: ‘Engaging the Ecosystem in Building a Resilient Approach to Cloud in Banking’ – join the session to hear real examples of the ecosystem working to de-risk third-and fourth-party consumption of cloud on 2 November.
- IBM watsonx - reach out to Casey.McDonald@IBM.com to request a demo on IBM watsonx and / or overview of the ISV Validation program.
- Join the newly launched Financial Services Cloud Council & Forum (FSCC&F), here. In FSCC&F you can find blogs, discussion documents, details of upcoming events, and whitepapers including the recently published Testing Operational Resilience whitepaper here.
September 2023 Presentations
#generative-ai#IBMwatsonx#watsonx.ai#watsonx.data#watsonx.governance#IBMCloud