How Workday has built a governance regime for Responsible AI
Our company was founded with the goal of inspiring a brighter workday for all. It’s in this spirit that we’ve been focused on ensuring that our AI ethics principles are fully operationalised across the company and aligned with the expectations of our customers, partners, and key stakeholders. We’re proud of our world-class data privacy program, and we’re applying the lessons learned in building and maintaining that program as we continue to expand our robust and customer-centric approach to responsible artificial intelligence (RAI) governance.
In our white paper, “Empowering Innovation through Responsible AI Governance,” we share in great detail what we’ve been doing with RAI governance at Workday. Below is an overview of the key takeaways from the paper, as well as what we have planned for the future.
Our Approach to Responsible AI
Our RAI program starts with a commitment at the very top. Key executives from across the company, including our chief integrity and compliance officer, chief diversity officer, and chief technology officer participate on our RAI Advisory Board, which is chaired by our General Counsel. This board meets regularly to review and approve new aspects of the RAI program and advise on novel issues as they arise.
In addition to the RAI Advisory Board, we’re growing our dedicated RAI team, which is responsible for programme design and administration, training and development, and orchestrating the ethical review of our AI and ML technologies. Our RAI scientists and technology experts pilot the development of RAI governance frameworks and bring items to the RAI Advisory Board for review and approval.
Our RAI team relies on input and participation from our network of RAI champions. These experts from across the organization assist in helping ensure a complete understanding of the technical, legal, and compliance details needed for governance, and serve as RAI ambassadors within their respective teams and as the local go-to resource for guidance and support.
Core Aspects of RAI Governance
We have built a set of RAI guidelines for AI and machine learning (ML) development at Workday which builds on recognised best practices, including those described in the NIST AI Risk Management Framework. The guidelines provide our product and technology teams with directions on how to implement our AI ethics principles in the development of responsible AI solutions.
One of the most important guidelines is to always incorporate human review of any outputs generated from the AI technology that we release. Our guidelines also highlight the importance of transparency and disclosure, which we operationalise primarily through our machine learning fact sheets, providing our customers with a clear understanding of how our AI and ML technologies are developed and assessed in order to help mitigate any risks associated with their use. Based on the specifics of the intended use case for the solution, additional guidelines in the areas of fairness and bias mitigation, explainability and interpretability, and maintenance and monitoring may be required, among others.
To implement and scale a risk-based approach to AI and ML development at Workday, we’ve created an RAI risk evaluation tool that our product managers (PMs) use when conceiving any new AI and ML project. The tool walks PMs through a series of questions to determine the sensitivity level of the technology and the appropriate set of RAI guidelines to highlight, relevant to the intended use case.
The Future of Responsible AI
Looking ahead, Workday is committed to further enhancing our RAI governance program. We continue to expand our RAI team and designate additional Workmates as RAI Champions, fostering a culture of awareness and accountability.
It’s also important to note that while we prioritize responsible AI at Workday, we also actively engage in advocating for it beyond our organization. We are vocal proponents of AI regulation that builds trust and promotes innovation. We actively contribute to AI-focused policy discussions around the world, including the US, the EU and the UK, to drive responsible AI practices internationally.
Workday is consistently recognised as one of the world’s most ethical companies, and we see our commitment to responsible AI as part of this. We always keep customer and stakeholder trust top of mind and work to ensure that our core values of integrity and innovation are reflected in the AI and ML solutions that we develop. By taking proactive measures, making thoughtful investments in this area, and engaging with valued stakeholders, we’ll help shape a future where RAI becomes a driving force for positive change at Workday and beyond.
You can hear more from Kelly at our upcoming Digital Ethics Summit (4 December) during the 'Meet the AI Ethicists' panel.
Authors
Dr Kelly Trindel
Dr Kelly Trindel serves as chief responsible AI officer at Workday.
Kelly directs Workday’s AI governance program, which is committed to crafting innovative and trustworthy AI solutions to solve real business problems.
As a pioneer in the responsible AI movement, Kelly has significantly contributed to the field, including testifying before the U.S. Equal Employment Opportunity Commission (EEOC) and later leading an EEOC task force on ethical AI—one of the government’s first. Her transition to the private sector in 2018 marked a continued focus in this area, including leading the establishment of ethical and workable guidelines and guardrails for applied AI development at both Workday and previous organizations. With more than 15 years of experience in quantitative science, civil rights, public policy, and AI ethics, Kelly’s influence and commitment to responsible AI are instrumental in driving the industry forward and fostering AI solutions that have a positive societal impact.
Kelly is a social scientist by training, having earned a Ph.D. in experimental psychology at the University of Texas at Arlington.
