How to Build a Business Case
Adopting Equitable AI Principles can make your organization a better and more inclusive place to work. It’s a strategic opportunity for an employer to be seen as a leader and innovator in its field.
We created the following guidelines to help organizations consider what can be included in your business case for Equitable AI. After reading these guidelines, refer to Play 1: Build an Equitable AI business case in our Equitable AI Playbook to learn valuable tips for building your own business case.
We derived these guidelines from an important and valuable resource, The Business Case for AI Ethics: Moving from Theory to Action, by All Tech is Human (CC BY-NC-SA 4.0).
Building a Diverse Workforce
Decades of research by organizational scientists, psychologists, sociologists, economists, and demographers shows that being around people who are different from us—including people with disabilities—makes us more creative, more diligent, and harder-working. Accenture has also found that companies who embrace best practices for employing and supporting more persons with disabilities in their workforce earn 28 percent higher revenue, double their net income, and have 30 percent higher economic profit margins than their peers.
Because technology underpins so much of today’s work environments, it’s essential that organizations consciously choose AI policies that support an inclusive, accessible environment where people from underrepresented groups can thrive.
Investing in a culture of responsibility, fairness, and inclusion across your organization can help grow its reputation and bring in top-tier talent. According to the 2019 Deloitte Global Millennial Survey, millennials and members of Generation Z gravitate to employers with organizational values and ethics that align with their personal beliefs.
Both consumers and employees have expressed significant concern about data usage, data protection, and privacy issues, including in hiring algorithms. The Pew Research Center found that 76 percent of Americans would not want to apply for a job if they knew they would be evaluated by a computer program or algorithm. Vetting the AI tools you use for candidate engagement will ensure that you test them for actual performance quality. Many AI tools on the market operate based on pseudoscience, and could easily overlook top talent.
Employees’ experience and institutional knowledge should be retained whenever possible. It is far more expensive to onboard new employees than it is to keep the talent you have. In fact, according to a report by the Job Accommodation Network (JAN), eliminating the costs of training a new employee is one of the primary benefits of implementing Equitable AI practices, providing employees with disabilities with the environment they need to be successful. However, employees may look for jobs elsewhere if they perceive privacy or surveillance risks from AI or if they believe certain AI-enabled decision-making could lead to inequities at work.
There is a compelling business case for ensuring the candidates you don’t hire leave with a positive experience. Some companies have found that the overlap between applicants and customers is significant. Virgin Media noticed a five million dollar revenue loss after six percent of their rejected applicants unsubscribed as customers.
Because people with disabilities are the third-largest market segment in the United States, businesses shouldn’t ignore their purchasing power. Applicants with disabilities who feel that they wasted significant time and energy only to be unfairly rejected by a bot may leave unhappy and frustrated.
Reduced Legal Risk
Technology often outpaces standards and regulations, though updated guidance could be imminent. Organizations that proactively create Equitable AI governance structures will have a significant advantage in the changing regulatory landscape. In addition to staying ahead of the regulatory curve, leaders who create internal guidelines and standards will also set themselves up as recognized experts, which carries another benefit. These organizations can help policymakers shape future regulatory processes that drive innovation by acting as industry advisors.
Enhanced Product Quality
Using technology in a socially responsible manner isn’t just the right thing to do—it will also help improve the product or service that your organization creates. That’s because you’ll become more likely to incorporate methods that perform best for all end users. In fact, that’s a major tenet of universal design. As with accessibility in general, IEEE’s Ethically Aligned Design Initiative stressed the need to create a sustainable culture of AI ethics in its call to action for businesses using AI (PDF).
Sustainability and Growth
Organizations that prioritize responsible AI practices now are better positioned for success in the future. Roughly 72 percent of Economist Intelligence Unit (EIU) survey respondents reported that ethical reviews in the development or use of AI are very important or critically important for shareholder and investor relations. Without being able to demonstrate clear practices to stakeholders and customers, an organization’s AI won’t inspire trust.
Organizations that adopt and promote Equitable AI will also place themselves in a better position to build ties with government agencies, institutions, and public sector entities that display a keen interest in responsible AI. Building these relationships as a trusted leader in Equitable AI could foster significant benefits through new contracts, grants, and relationships.
When designing Equitable AI systems, it’s useful to make the underlying processes, code, and algorithms as transparent as possible. It’s also important to clearly document the data sources used to train AI models. Having transparent, well-written and well-designed AI helps people working with AI systems to grasp the reasoning behind predictions and decisions, regardless of whether a person or AI tool led the recommendation. This in turn facilitates effective, and equitable, long-term operations. By building organizational understanding of how products and processes function, transparent AI enables efficient feedback loops for continuous improvement.
Organizations using AI should also be transparent—both internally and externally—about when and how they are using AI. Transparency about the use of AI, coupled with transparency about the design of AI, will enable different stakeholders to participate in the development process. This means diverse teams, including people with disabilities, can help ensure AI systems are both transparent and equitable, which in turn helps ensure the AI’s sustainability and the organization’s growth.
The importance of data and analytics to business growth and digital transformation will only grow. Safeguarding privacy and ensuring the accuracy of data is essential. Aside from the significant cost of data breaches, any data being collected now through AI technologies is also a highly valuable commodity for an organization. Putting policies in place is essential to ensure that the data repository you create is as accurate as possible and that it is safeguarded against cyberattacks.
This Business Case for Equitable AI is based on the following resources:
- Staying ahead of the curve: The business case for responsible AI, Economist Intelligence Unit (EIU)
- The Business Case for AI Ethics: Moving from Theory to Action, All Tech is Human
- IEEE Ethically Aligned Design: A call to action for businesses using AI (PDF)