Financial Services firms turn to data ethics to manage digital risks

In 2021, data ethics will be the tool financial services firms choose to manage digital risks in automation and cybersecurity.

In 2019, the California-based lender loanDepot made headlines when it managed to offer an end-to-end digital mortgage that closed in just 8 days, more than 80 percent faster than the 45-day industry average.1

Digital solutions enable this kind of agility at scale, but individual and societal risks can also scale in digital systems. For instance, if the training of the loan approval model utilized historical data that exhibited lower loan amounts for women and minority groups, the risk of that discrimination being carried forward is very real. Practicing data ethics helps organizations to identify these risks early in the development of new products and services, and to intervene with tools, assessments, and governance processes. This allows them to act before products and services are ever deployed, mitigating the risks that could have a material impact on peoples’ lives.

A 2020 Avanade Digital Ethics Study2 found that 70 percent of financial services firms are planning to increase their investment in data ethics over the next 12 – 24 months. These findings mean that the same industry that spends an average of $2,700 per employee annually on cybersecurity (more the twice the cross-industry norm) is realizing data security measures are only part of the solution to address growing digital risks.3, 4 They also reinforce what Accenture is seeing in the financial services industry: organizations are aligning on mature ways to address systemic data ethics risks, moving on from bias and fairness assessments for newly automated systems, which proved insufficient at managing digital and ethical risks.

The financial services sector looked first to adopt technical solutions, some quite innovative. Capital One, for instance, has credited the use of synthetic data for improving both fraud detection and privacy.5 But just as Model Cards, a practice of tracking features and risks of AI models, might be an early step on a journey to actively manage deployed AI models, fairness assessments are often an early step on a bank’s data ethics journey. Soon after exploration of technical tooling, the need to take additional steps to mitigate digital risks becomes apparent. These steps frequently include modifications to organizational culture, business strategy, and stakeholder engagement.

Other important steps on the data ethics journey are a thorough exploration of the ethical considerations of data analysis and use, and then an examination of data provenance and management. However, the firms willing to do the hard work of embedding data ethics into their culture are the ones that will enjoy sustained progress in this space and a market-differentiating ability to manage potentially out-sized risk.

With an increased use of artificial intelligence (AI) as well as the more frequent publication of ethical guidance by regulators, the C-suites of our large financial services clients are concerned about the potential ethical implications of data-centric use cases more generally. These concerns are increasingly common and are prompting a renewed interest in how the organization’s core values show up in products and services. Specifically, financial services firms are seeking to understand how their data processing aligns to the firm’s core values and the steps which can be taken to ensure that alignment is propagated throughout the firm.

Core values indicate what firms care about and want to protect and promote. Often, principles – which generally describe how to implement the values – are derived from the core values. And sometimes, codes of ethics are built from the principles for specific use cases, products, services, or departments. Governance then assesses whether the norms established from the values, principles, and codes are satisfied in particular cases. Ethics is the work that is done to satisfy the values, in accordance with the principles/codes, and in support of governance.

For example, Accenture’s Risk, Applied Intelligence, and Responsible Innovation teams have collaborated to help a client define a clear vision for a leading data ethics program that leverages not only leading fairness and privacy tooling, but modifies governance, controls, and training to include data ethics considerations and a roadmap to operationalize it. This work, which is now being operationalized, is grounded in a set of principles and best-practice guidelines that define ethical use of data across the firm, as well as governance controls and a framework that support effective monitoring and oversight.

Now, instead of advanced technologies introducing new risks and liability, incorporating data ethics proactively helps to mitigate such risk. What’s more, it enables teams to use AI as an enabler of new revenue opportunities and competitive differentiation. Strong data ethics practices allow organizations to grow without being encumbered by the shifting landscape of privacy laws. Going beyond compliance helps financial services firms to stay ahead of the competition and erect barriers to entry for new competitors.

With clear strategies to manage ethical risks across innovation and change initiatives, our client is better able to reduce digital risk exposure and protect stakeholder interests. In doing so, it is well-positioned to talk about what it’s doing in a way that will help to build trust in its brand, which adds gravity to attracting and retaining customers.

While this client is now a leader in its market, its challenges and journey are far from the exception. These days, lenders, insurers, and capital markets firms must have technological chops on par with the most innovative platform companies in order to compete.

As part of operationalizing data ethics, some organizations employ Accenture’s Ethical Spectrums framework to offer scaled decision-making with increased agency because our research shows that companies that scale AI successfully achieve 3x higher returns on their AI investments.6 In other cases, attention is focused on holistic training programs that improve critical thinking and connections with concepts of justice and human rights in the development of AI systems. Regardless of the approach, with their robust data ethics strategy in hand, business leaders can be sure they are managing digital risk. And beyond that, organizations can now responsibly carry out successful future strategies – for revenue, growth, innovation, stakeholder engagement, and digital competitiveness and differentiation – unencumbered by a constantly shifting regulatory landscape.

With a strong data ethics strategy and industry-leading practices in place, financial services firms are well-poised to compete in today’s agile marketplace and be responsive to the digital native customers that will become the majority of their revenue in the not-too-distant future.

  1. “New Rapid Mortgages Allow Closings in as Few as 8 Days”, March 13, 2019.
  2. “The Good, the Bad and the Ethically Responsible: Consequences of AI”, Avanade, August 13, 2020.
  3. “Financial Firms’ Cybersecurity Spending Jumps 15%, Survey Finds”, Bloomberg, August 4, 2020.
  4. “Three Benchmarks to Inform Cyber Security Spending Plans for 2020”,, November 29, 2019.
  5. “Why You Don’t Necessarily Need Data for Data Science”, Capital One, 2019.
  6. “AI: Built to Scale”, Accenture, 2020.

Source link