Mastercard's Practical Implementation of Artificial Intelligence Oversight
In the ever-evolving world of artificial intelligence (AI), Mastercard's AI governance team is making significant strides in ensuring that AI remains effective, inclusive, and trusted. Led by John Hearty, the Global Lead of AI Governance, the team is responsible for mitigating bias, ensuring efficacy, and enabling transparency in AI.
Hearty's team at Mastercard has been actively collaborating with peers, including developer teams, to review the company's proposed AI governance framework. They have also worked together on bias testing an API, which will be distributed within the company. This collaboration is crucial in creating a robust and reliable AI system.
The team's efforts extend beyond Mastercard's boundaries, as they have partnered with a data science team to work on a model documentation template. This collaborative approach is essential in fostering a culture of transparency and accountability in AI development.
In 2022, Hearty's team at Mastercard created an AI governance strategy aimed at building influence and making compliance easy. This strategy has been instrumental in managing the increasing number of AI systems that the team needs to assess, give guidance on, and monitor. The number of such systems has doubled every year up until 2024, when it increased by 60%.
Small teams can have a significant impact in AI governance despite the talent shortage. The team at Mastercard, comprising an architect, model risk expert, technology specialist, communications expert, and a lead with a background in R&D leadership in data science, is a testament to this.
One of the challenges faced by the team is access to qualified AI governance talent and skills. However, they have managed to navigate this challenge by partnering with developers to create an API for bias testing and mitigation requirements for large language model evaluation.
The IAPP finds that 90% of organizations using AI are working on AI governance. Hearty acknowledges the importance of this, stating that 61% of people are wary of trusting AI and 67% report only low to moderate acceptance of it. Pre-contract information sharing helps create trust, and the team at Mastercard has implemented this practice.
Operationalising AI with a small team is practical if influence is managed well, a resilient culture is implemented, smart processes are refined, and enablement is focused on. The team at Mastercard has shown that this approach can lead to significant progress in AI governance.
However, the team also faces challenges with record keeping and documentation, as different teams may not document, monitor, or own their solutions in the same ways. This is an ongoing challenge that the team is working to address.
Despite these challenges, the best partnerships are structured to provide value to both parties. The team at Mastercard continues to forge such partnerships, ensuring that they contribute to the broader goal of making AI more effective, inclusive, and trusted.
Read also:
- Industrial robots in China are being installed at a faster rate than in both the United States and the European Union, as the global market for these robots faces a downturn.
- Hyundai N affirms transition to hybrid performance-centric models, initiating with Tucson N
- EAFO Research Uncovers Crucial Elements in Electric Vehicle Adoption within the EU
- Stock markets in India anticipated a moderate opening, influenced by mixed signals from global markets.
