HR and AI - how Mastercard credits AI as an enabling technology, delivered with due governance

AI is already an essential technology in the financial services industry. Mastercard is a useful case in point. The firm processes around 110 billion transactions each year, and uses AI cybersecurity tools to monitor and detect potential areas of fraud. So perhaps it is not surprising that the organization is keen to use the latest AI tools throughout the company, including in HR. 

Speaking at a recent event hosted by the Josh Bersin Company, Michael Fraccaro, Chief People Officer at Mastercard, said one of the financial services giant's earliest HR applications was to help with diversity inclusion in recruitment:

We used a platform tool that was gamifying particular tasks or activities, particularly for graduate recruitment. It did not ask any biographical details, did not ask what university you went to, what your gender, race, or ethnicity was. It didn't ask any of those questions, it just wanted to know, once the candidates went through this platform, whether they could do the task. So it was the first screening approach that we had to bring them into the organization for the next round of interviews, and that did reduce the level of potential unconscious bias.

Having successfully used AI at this level, last year the firm introduced an automated scheduler, which has increased productivity by 87%.  It helps staff look at a range of opportunities to meet with their managers, and be able to schedule these dates without having a human interaction, he explained: 

It goes through everyone's calendars, say there are five different people's calendars and tries to make an appointment. This scheduler has AI built into it that enables this task to be simplified, and much more efficient. And if there is a change, which likely there are changes to schedules, this allows them to be able to do that too.

Mastercard also now uses an AI based platform that matches skills and experience with projects and jobs.  Fraccaro said:

We call it an internal talent marketplace. On one side of the marketplace is the employee who will go into this platform, identify their skills, their aspirations, their experiences, and what they're looking for. And on the other side, you have managers who put a project or an open position onto the platform.

The tool will match the preferences of the individuals with the available projects that are on the platform. This has produced clear deliverables, according to Fraccaro: 

We've seen an increase in the number of people getting matched to projects and what I love about this particular platform is that it enables a much more democratized approach to talent management as well, because you're not necessarily having to recruit from the same business vertical. If the individual has got the skills and requisite experience to be able to do a particular project, they will get matched to it.

The tool has also increased the number of people being teed up for mentoring. It matches interests and individuals between what they're looking for from a mentoring relationship to the mentors that are available on the system.  Fraccaro said:

We've seen 1000s of matches globally around that. And the more people use it, the better It becomes.

The next version will integrate learning, training and development opportunities as well, he revealed:  

So if someone has a particular training need, the tool will begin to be able to direct specific kinds of training solutions. I think the big thing for us with all these platforms and tools is that they are going to get better and better over time.

Governance

But he warned that governance of these AI-based tools has to be a priority. Mastercard has an AI governance steering committee that looks at particular cases to make sure that governing principles are followed. There is a focus on trust, data, and data privacy, and security, as well  as who owns the data. 

This committee contains business leaders from throughout the company, including Chief Data Officer, people from the compliance team, and representatives from each of the critical business and functional areas.  Fraccaro commented:

From a commercial perspective we've got to make sure that we maintain trust with the customers. Having governing principles really ensures that we're making the right decision in the way that we're implementing and monitoring the various tools that we have.

He added:

I think anyone that's embarking upon this journey needs to make sure that they've got this governance or these guardrails to ensure it's done in a very thoughtful way.

For example, when ChatGBT was launched the company knew that employees would be curious and look at how they could use it.  Fraccaro recalled:

We were very clear in terms of some guidelines to say, Hey, this is okay to play with it. It's still an experimental phase. But here's some things that you should avoid and don't do. So for example, putting together a client brief or, or using it in the way of interacting with a particular client. There are some guardrails, because at the moment, it is still nascent. These are just things that you need to be aware of, but we don't want to restrict the curiosity of our employees either. 

Going forward, he cautioned that in the end there is still a human on the other side of the response:

Whether it's the recruiter, or the hiring manager, they still are accountable for what goes out there to the market. We do not want to see a generative AI job that goes out there without the verification of someone.

He believes that a tool like ChatGBT will take out the drudgery of some of the tasks that people do, and that could create 40% more time during the day to have strategic conversations, and more deep and meaningful conversations with candidates or clients.  But this does lead to the biggest question - ‘is my job safe’.  Fraccaro argued:

I think there's a whole cycle that we see with any new technology being introduced. The fear, then denial, and then finally getting to change and acceptance. So, embrace this change, because it is happening, and the more that you can lean in and understand what it is, and what it isn't, actually will help individuals longer term. 

Mastercard experienced this at its service center a few years ago, as it started to put in place some automation tools, he noted: 

We did not lay people off, we retrained them and rescaled them to train these bots to be able to do some of these automated tasks, and it freed them up to do other things as well. The up-skilling, re-skilling and the re-tooling is a real opportunity.

Overall,  Fraccaro believes that AI is just another tool, although a very powerful one:

It's part of evolution and, but it's not going to change the way that humans are motivated by their emotional state. So you know, if you're engaging with your CEO, he's not going to want a ChatGBT response; he's going to want a heartfelt one connecting to what he or she is feeling. Those aspects are really important.