AI & Ethics: The Grand Challenge for Our Generation

By ibmblogsMay 4, 2021

Young Asian woman with protective face mask unable to unlock her smartphone with face recognition software in the city. Biometric verification and artificial intelligence concept

Building a Roadmap of Trusted Data & AI for Financial Services

What do nuclear proliferation, climate change and artificial intelligence (AI) have in common? All three, left unfettered, have the power to negatively impact the world as we know it.

AI is rapidly emerging as an important tool for solving humanity’s challenges, but it is not without its risks. People are increasingly uneasy about allowing intelligent and autonomous machines to access their personal data to make decisions. This has prompted a growing demand for ethically aligned AI that creates trust and protects privacy, fends off would-be hackers, and prevents bias.

I believe trusted data and ethically aligned AI is one of the “grand challenges” of our generation. Developing and enforcing standards, ethics and governance around the use and development of trustworthy AI systems has become an urgent priority.

“Like all disruptive technology, artificial intelligence offers both great promise and the risk of great peril; and the increasing sophistication of AI has rightly sparked a vibrant debate on how best to balance the two. As AI systems become more ubiquitous, trust in these systems will be critical and organizations will need to demonstrate that they have thoughtfully considered the unique risks of AI and responded in a manner that is consistent with their corporate values and aligned with the expectations of clients, employees, and society.” William Stewart, Head of Data Use and Product Management, Data and Analytics (DNA), Royal Bank of Canada (RBC)

The European Union (EU) unveiled strict horizontal regulations and first-of-its-kind policy on April 21, 2021 to govern the use of AI. The draft proposal would impact use cases like self-driving cars, hiring decisions, bank lending, school enrollment selections, scoring of exams, and more. It would also affect the use of AI by law enforcement and court systems — areas considered “high risk” because they could threaten people’s safety or fundamental rights. Companies that violate the new regulations could face fines of up to six percent of global sales.

Private sector and financial services (FS) leaders across the globe view ethically aligned AI as an integral part of business operations, with the board of directors playing a critical role in establishing and overseeing its adoption and implementation. It is predicted that by 2023, AI regulation for FS will be in effect in places like Singapore, Europe, UK, Canada and Australia.

Clearly, it is no longer a matter of if AI will be regulated, but when. So, what are the next steps toward ethically aligned AI in FS to prepare for these regulations?

Banking on Trusted Data & AI

In today’s competitive environment, FS institutions are using data – including customer, firm, and third-party data – to deliver innovative alternatives to traditional banking solutions. With open and trusted data as the main currency, open platforms deliver relationship-based services built on customer engagement and experience and in so doing, allows banks and the emerging fintech companies to leverage their complementary strengths, instead of competing.

Driven by trusted digital relationships between banks and customers, this new business model is powered by AI and generates value even during a pandemic lockdown.

However, many customers are rightly concerned about how and when their personal data is being used. Trust in AI systems will increase only when customers are confident that their data and insights are protected, not leveraged for profit without their permission.

“Data ethics aren’t a cost of doing business, they are an investment in good business. We have a responsibility to set ethical standards that ensure transparency, minimize biases, and encourage accountability.” Terry Hickey, SVP & CIO Enterprise Data, CIBC

Customers and investors will flock to the most trusted brands in AI – and financial institutions who take a leadership role will benefit the most in the platform economy. Consider:

According to a recent IBM Institute for Business Value survey, more than 68 percent of consumers are willing to share personal information and data with their bank or other financial services institution. Almost 91 percent of individuals who share personal data with their bank trust it to protect their personal information and data to at least a moderate extent, second only to their own employer.
AI hyper personalization increases the likelihood of attracting, retaining, and delighting customers with a company’s product offerings, services and 24/7 availability.
Data and AI also provide new tools in the fight against fraud. When effectively used, these tools improve fraud detection, reduce credit losses, and enable know-your-customer regulatory checks.
The future relationship between bank and customer is likely to be much deeper and much more interactive. But getting there from here requires radical transformation across business and operating models, as well as changes in the way resources, business processes, and technologies are assembled to create value.

Creating the IEEE Finance Playbook

While most FS organizations agree in principle that ethically aligned AI is a necessity, they do not have the resources to build, execute, and scale trusted data and AI applications in isolation. Building a solid foundation of trusted AI can only happen if government, industry, and standards organizations work together.

To respond to that challenge, IBM joined forces with a global ecosystem of partners in 2019 to create an AI standards guide focused on ethically aligned AI for financial services. We worked in collaboration with all the Canadian major banks, pension funds, as well as the thought leaders at IEEE, the world’s largest technical professional organization devoted to advancing technology for humanity and creators of the globally recognized document, Ethically Aligned Design (EAD).

Based on EAD principles, we created the IEEE Finance Playbook v1.0: Trusted Data & AI Systems (AIS) for Financial Services, which harnessed the cutting-edge thinking of dozens of financial services AI experts. It is a significant step toward developing FS standards and certifications, while nurturing fintech ecosystems and hyper-scaler alliances.

I am privileged to chair the Industry Executive Steering Committee that oversaw the creation of the Playbook, the survey (providing research for future iterations of the Playbook) and continues to plan executive workshops and other outputs. For this work, we have been able to curate the implementation of best practices and roadmaps from the financial services industry, academia, NGOs and standards organizations. All six Canadian major banks, as well as credit unions, pension funds, and fintechs participated in this initiative.

“We are in the business of trust. A primary goal of financial services organizations is to use client/member data to generate new products and services that deliver value. Best in class guidance assembled from industry experts in IEEE’s Finance Playbook addresses emerging risks such as bias, fairness, explainability, and privacy in our data and algorithms to inform smarter business decisions and uphold that trust.” Sami Ahmed, SVP Data and Advanced Analytics (DNA), OMERS

Five key takeaways from the IEEE Finance Playbook are:

A first step towards successfully implementing trusted AI is developing standards and certifications, governance and reporting while nurturing fintech ecosystems and alliances.
Ethically aligned AI best practices that are emerging globally, useful for building a trusted roadmap that engages all lines of business across the C-suite.
Twenty high-value AI use cases that demonstrate how trusted data and AI can translate into business value in areas such as: engaged customers, increased wallet share, customer acquisition, reduced credit loss, and improved profitability.
A must-have framework to develop a global financial services community with ongoing surveys and up-to-date POVs.
Recommended key resources to help develop the three critical building blocks of Trusted Data and AI (people, process, technology) that allow organizations to deliver on the high-value AI uses cases.

The Roadmap to Trusted AI

The march toward AI regulation is inevitable. A commitment to ethically aligned AI reassures customers, partners, and employees that their privacy is protected, and their data is used responsibly.

“We are at a critical junction of industrial scale of AI adoption and acceleration. This IEEE Finance Playbook is a milestone achievement and provided a much-needed practical roadmap for organizations globally to develop their trusted data and ethical AI systems.” Amy Shi-Nash, Global Head of Analytics & Data Science, HSBC

IBM is committed to leading the way when it comes to solving the big issues surrounding trusted data and ethically aligned AI. For more than a century, we have been creating solutions that build collaboration and trust between people, businesses, and the machines they use. Our three principles for Trust and Transparency for AI are at the core of everything we do.

IBM Canada is taking a leadership role with IEEE by building community and establishing relationships with experts and targeted buyers – building on the fact that Canada is known and trusted for its global research and leadership in both AI development and financial services regulation.

Your company can benefit from being part of the journey toward ethically aligned AI by:

Shaping global AI regulation standards by participating and contributing thought leadership to IEEE Finance Playbook and Survey. Download the Playbook here for information on how to participate.
Building the trusted brand. Customers are less likely to use a banking app or offering if they don’t trust the data or the AI systems.
Appealing to top talent, particularly millennials and younger, who are drawn to companies that use data where AI is being designed and implemented in ethically aligned ways.
Attracting investors by highlighting your contribution to the UN’s Sustainable Development Goals, particularly the guiding principles for Open Banking and ESG Transformation.

“In financial services, deployment of AI technologies is both an opportunity and a responsibility. It can positively transform client and employee experience, and augment core capabilities of organizations. We also have a key role to play in deploying AI technologies responsibly and ethically, in order to maintain trust of clients and other stakeholders. IEEE has successfully brought the ecosystem together to tackle important questions, an essential component in helping the financial industry continue its journey to set exemplary standards for responsible and ethical AI.” Mathieu Avon, VP, Integrated Risk Management, National Bank of Canada

We Invite Your Feedback and Participation

Get in on the conversation! We invite you to benefit from the IEEE Finance Playbook ecosystem. Take part in the standard setting discussions, shape industry policies, and influence trust and brand value of the financial services industry globally.

Suggested next steps:

Download the Playbook here. Use it as an internal resource, or as a door-opener for C-level discussions on ethically aligned AI.
Take the 20-minute companion survey on Trusted Data and AIS Readiness and receive an evaluation of your organization’s trusted data and AIS readiness and guidance on how best to use the playbook.
Engage with us. Contact IBM experts and our clients to do an executive briefing, workshop, co-publish a POV, or to be part of a conference panel.
Email me to share your feedback: pavel@ca.ibm.com

Suggested further reading

Essentials for the post-pandemic bank

Banking on the Platform Economy

Advancing AI ethics beyond compliance