In an exclusive conversation with MartechAi.com, Mukta Arya, Chief Human Resources Officer (Asia-Pacific) at Société Générale, discusses how artificial intelligence is reshaping the world of human resources. Speaking with Editorial Lead Brij Pahwa, Arya reflects on her three-decade HR journey from the days of paper résumés to predictive analytics and internal GPT tools. She shares insights on how AI is transforming recruitment, learning, and workforce analytics while stressing the continuing importance of human intuition, empathy, and storytelling in the age of intelligent automation.
Question: As CHRO of one of the largest European banks in the APAC region, you have seen how talent management and HR have evolved. There was a time when you had to sort through paper resumes. Now everything is digital. You used to receive only a dozen or two CVs; now thousands pour in each day. You have seen this transformation over decades. How would you describe that change? When did the idea of AI first enter HR, and what does it do in HR today?
Answer: I started my HR career in 1997, when even an MBA specialization in HR was still called "Personnel and IR," so you can imagine how things have changed. At my first job, all our records were physical files and we had paper resumes. Even then, we had an HRIS (Human Resources Information System) at my company and it was quite advanced for those days.
But we never thought about technology as advanced as AI back then. Over the years, with the rise of the internet and various online recruitment tools, things improved a bit, but now it is a completely different ballgame. For us, as an international bank, talk about AI began a few years ago, and we even created a group called SoGen AI to explore AI use cases for our daily banking operations.
In HR specifically, since many companies started offering AI tools, we began experimenting with them. For example, we tried AI for learning content curation, candidate sourcing, and chatbots. What started as pilot projects is now embedded in our systems. For instance, our learning management system uses AI: based on an employee’s interests and learning history, the AI suggests training programs that are suitable for them.
We also have an internal talent marketplace called ACE: it is like an internal LinkedIn. Employees create profiles, and the system matches them to relevant positions based on their expertise. These are ways AI is already established in our HR processes.
We also use AI for predictive analytics. Right now it is very basic, for example, trying to predict which employees might leave. It is at an early stage, but slowly it is becoming more effective. I believe AI will reign supreme in HR in the next few years. People with AI skills will help ensure our HR processes include AI elements without compromising quality or introducing bias.
We even have our own internal GPT instance, and we call it SoGPT, and we use Microsoft Copilot internally. As a bank, we must be careful about data privacy and confidentiality, so we cannot use external AI tools. With our internal tools, for example, we have employee satisfaction surveys that are about 50 pages long. Using Copilot, we can get those reports summarized in minutes. We are already using our internal GPT and Copilot to help interpret various labor laws, for instance.
Overall, AI is being used in several areas of HR for us already. We still have to wait and see the major impact, probably in the next year or two.
Question: You once said HR professionals are “artists with a scientific mind” who turn workforce data into strategic insights. Can you give an example of a recent instance where data led to a significant people or HR decision? Did it surprise you?
Answer: It happens all the time. HR has a huge amount of data, but data alone has no meaning unless you weave it into a story that shows impact.
For example, we looked at our benefits usage. A simple case: we ask, where are employees using the benefits the most? That tells us what they really need. If some benefits are hardly used, further increase of it may not bring us proportionate benefits.
Data can really surprise us. Another area we analyzed was well-being. We looked at satisfaction survey responses where people reported being stressed or fatigued, and we combined that with operational data, like who was not taking vacations, where managers were frequently declining leave requests, and who was using a lot of sick leave and medical benefits. Of course all this data is anonymous, but when we saw patterns, say, an employee not taking any vacation, having leaves denied, and then taking many sick days, we knew something was wrong. By overlapping these factors (like a Venn diagram), the intersection revealed a problem. We identified a few such cases and flagged them to the managers or HR business partners in those areas, so they could check in on those employees. That is a case where data clearly told a story we might have missed otherwise.
Data helps bust myths too, when it is communicated properly. There is always some belief in the Financial industry that it is hard for female to progress in the corporate ladders and banks are not doing enough to improve that. By analysing the trends of demographics of the senior management in the bank, we could observe the positive change we made over the years and we could see more females taking on senior roles. More importantly, we could then communicate these achievement with the wider staff community so that we could bust the myths by facts with transparency.
These examples show how assumptions can be wrong if you do not check and systematically report the data. As HR, our role is to be custodians of this data and turn it into informed decisions. Based on these insights, like whether to invest in an internal mobility program or how to tailor benefits, we make strategic choices. In short, all important HR decisions should be driven by data.
Question: HR has traditionally been human-led and intuitive. Now AI can evaluate performance and predict attrition. Do you think HR leaders risk outsourcing judgment to algorithms? Where should the line be drawn between data-driven decisions and human intuition?
Answer: To me, it is a combination of both. HR is really an art and a science. You need the science, the data, and the art, human intuition and context. You cannot just follow numbers blindly. Context matters, and there will always be intangibles that data alone cannot capture. I believe even years from now, HR’s role will be to provide that intuitive touch. AI might learn context, but we will still need people to decide when and how to apply insights, so that decisions remain balanced.
In other words, it is a collaboration. AI tools can do some tasks faster and more creatively than humans, but we still need people to put the whole picture together. We still need HR professionals to interpret and act on the data. It is not one or the other, the best results come from working together, using AI for what it does well, and using human judgment for the rest.
Question: There has been controversy about using AI to shortlist CVs. With so many applications, companies use automated filters to screen candidates. But that raises the risk of filtering out excellent candidates just because their CV did not match certain criteria. What checks do you have in place to make sure good candidates are not missed? After all, some professionals (like journalists) use non-standard CV formats.
Answer: There is always that risk even with manually screening of CVs, where we just keep the criteria in our mind and even human judgement may not be fair at all times. When you have thousands of applications, efficiency matters and it is not humanly possible to screen them all manually. So I understand some level of automated filtering could make sense. Candidates know how to optimize their CVs to get past filters, and employers should also ensure our filters are not biased against any group.
As of now, from what I could see from the available solutions in the market, I do not have a perfect answer, but it is an issue employers should actively working on. We recognize the need for checks and balances and continue to look for ways to mitigate this risk. Of course, CV shortlisting is only one part of the recruitment and selection process, where there are other “people” element like interview to ensure candidate suitability and there are also other means to ensure the access to our target candidates, like through employee referral and headhunters
Question: As AI automates tasks like content writing or planning, are managers starting to expect smaller teams? In general, have you seen AI reducing jobs or creating new ones?
Answer: Right now, I see this as a transformation phase. Everyone is still figuring out what AI can and cannot do. I agree that in the future, some roles could change completely and certain tasks might be automated, but we will also need people who understand AI to implement and manage it. At present, I have not seen a big push to drastically shrink teams because of AI, it is early days. We are still experimenting with use cases and tools.
So far, I have not seen any major headcount cuts directly due to AI. Our industry (corporate and investment banking) is still very people-intensive. But I do expect changes in the next few years as more processes become automated. For now, AI is reshaping roles rather than eliminating them outright.
Question: Speaking more broadly, talking with other HR leaders, have you noticed companies cutting jobs because of AI? I saw a chart where job opportunities declined sharply after large language models emerged. Is that trend playing out in HR?
Answer: The trend of AI impacting jobs is certainly happening in some activities. Tasks like KYC checks or basic auditing, for instance, are increasingly automated. In HR, we are beginning to see signs too.
I think eventually some jobs will be affected in one way or another. For example, if chatbots handle all first-level enquiries and request from staff, you might not need as many people for those tasks. But then you will need people to program and oversee those chatbots instead. In our own company, we are talking about where AI can add value without sacrificing quality.
Right now, I would say we are still in a transition. The chart you mentioned might reflect where things are headed, but currently in HR we still rely heavily on human expertise. I believe that in a couple of years, AI will definitely lead to some changes in staffing. But today we are mostly evaluating, learning and progressively adapting.
Question: Many interesting examples. Just before we wrap up, thinking about your own career and legacy, what kind of HR legacy would you like to leave in the AI era? What do you hope future leaders will build on?
Answer: That is a great question. Ideally, I would like to see solution-driven AI integrated across the entire employee lifecycle, from attraction to exit. For example, having tools or apps that gauge employee mood or engagement at onboarding and regularly thereafter, complementing human check-ins.
Ultimately, I want to see HR and AI coexisting seamlessly. That means building an HR team that is well-versed in data and AI, knowing how and when to use these tools effectively, and when a human touch is needed. There should not be one-size-fits-all solutions. The key is having the sensitivity to apply AI in the right places, while preserving the human perspective where necessary.
So, my legacy would be an HR organization where people are trained in data analytics and AI, using them together to tell the story behind the numbers. Because no matter how advanced AI gets, decision-making will always need someone to connect the dots and add that human perspective.
Related Articles
