In today’s rapidly evolving business landscape, the integration of Generative Artificial Intelligence (Gen AI) is poised to redefine the way organizations operate and innovate. As this transformative technology becomes increasingly integrated into enterprise practices, ethical considerations take center stage, requiring organizations to navigate uncharted territory and address pressing dilemmas. In this week’s CDO insights we have with us Sid Banerjee, Chief Executive Officer at SG Analytics a research and analytics firm, where he is committed to leading the company towards an IPO. His career includes leadership roles at BRIDGEi2i Analytics Solutions, Deloitte India, Credit Suisse, Markit, and Deutsche Bank, where he has contributed significantly to business development, strategy, and operational meetings. Sid brings a wealth of experience to the field of data analytics and business leadership.

Sid Banerjee’s forthcoming discussion will address key facets of Gen AI. He will examine the evolving landscape and its impact on business evolution. Sid will focus on the pressing ethical concerns in AI implementation, methods to ensure transparency and accountability, and SG Analytics’ approach to mitigate bias and ensure ethical AI use. The conversation will also explore data protection measures and the role of data science service providers in Gen AI’s ethical development.

AIM: Could you help us comprehend the consequences of enterprises failing to embrace Gen AI, given that successful organizations historically owe their resilience to their ability to evolve? Why is the adoption of Gen AI imperative for organizations in today’s rapidly changing landscape?

Sid Banerjee: Gen AI has many connotations. Some people almost think it’s like a silver bullet, that it will solve every problem they have. Equally, some people are so skeptical about it, thinking it’s just a phenomenon that will come and go, leaving them untouched. In my opinion, both views are wrong. It will have its usage and a huge impact on the way people work. It will force people to reinvent themselves. But at the same time, it won’t solve every problem, nor will it fail to address many issues we have today. All its answers are going to be flawed. So, we are somewhere in the middle where it’s going to be a “horses for courses” strategy that people will follow when it comes to Gen AI.

Speaking of organizations, I’m reminded of Marshall Goldsmith’s book, which talks about what got you here won’t get you there. Every organization will have to think about it like that. What they have been practicing to do a bunch of things tomorrow may have to be done differently because the client’s expectations are going to change, as will the expectations of every stakeholder. The way people ingest data is also going to change, and in that evolving dynamic, it’s going to force almost every organization to take it very seriously and go down this path. But how it will impact each of them is something that all of us will see in the next 12 months.

AIM: Now that we’ve recognized the inevitability and significance of Gen AI adoption for organizations, particularly larger and successful enterprises, what ethical considerations loom large in the context of Gen AI? This becomes even more pertinent as many uncertainties still surround this evolving technology. 

Sid Banerjee: There is no one size fits all for Gen AI. It will vary not just for the specific industries that we are talking about. It will also depend on which functions or horizontals of that industry are going to deploy Gen AI. An example is within banking, The way a risk function will look at Gen AI will be very different from the way sales and trading will look at Gen AI. Equally, the way banks will look at Gen AI will be very different from how a media company will look at Gen AI. So the use cases are phenomenal. The opportunities of Gen AI could also be exponential. But it is like one of those toys that we have played with or the tools that you use, which if not rightly used, can lead to a lot of disturbance and disruption in the system that may almost force you to go back twenty steps instead of going forward 10 steps. So it needs careful consideration before it is implemented. And it also needs very thought through application of Gen AI both from an use case standpoint as well as its output and delivery aspects.

AIM: When it comes to the ethics of technology, transparency is a critical aspect, particularly in areas of high concern or with high-stake decision-making. What steps should organizations take to ensure transparency in their Gen AI applications or services?

Sid Banerjee:  You make a very good point in terms of the ethical concerns, very relevant also from an SGA standpoint because we give it paramount importance in whatever we do, as we believe that life is possible. From that context, I would say the three things that really worry people dealing with Generative AI, and more importantly, those deploying Gen AI are: Is it fair? Is it expandable or scalable? And lastly, is it reliable?

Why are these three important? Because when a lot of this is done by the machine and using terabytes of data from all sources, it becomes very challenging for anybody using this to be comfortable with these three points. The fairness obviously comes because a lot of this is dependent on historical data. And when you deal with historical data, many times the world has moved, circumstances have changed, parameters have changed. Therefore, what has happened in the past will not necessarily be happening in the future. It is not going to be reflective of the past so the patterns. That needs to be considered.

Equally, when we talk about expandable or scalability, many of them, you’re looking at a dataset which may not be a good representation of the rest of the dataset. And when it talks about reliability, then it’s a question of how the data has been interpreted by the system. Interpretation is, again, of big importance. So this is where I feel there is a need to be very aware of this and therefore bring in an element of curation of these models by human interface and not just deploy exactly what is being thrown out by the machine. And that is a sweet spot, which I think as a firm we are trying to address.

AIM: How is SG Analytics approaching the ethical challenges of Gen AI, especially when working with large enterprises across sectors like retail, pharma, healthcare, and banking?

Sid Banerjee: Let me share with you what I exactly see in this space in my conversations with CXOs of large organizations. There is a terrific pressure coming to the organization’s CEOs and CXOs from their board, where they want to see the benefits of Gen AI. They want the organization to adopt it; they are hearing a lot about it and they’re asking the question on how these organizations are embracing it.

But equally, when people at the leadership level are talking about Gen AI, they are themselves very careful of ensuring that they do not go down the wrong path in terms of deploying Gen AI because it can get you a lot of bad press in the event it goes off the handle.

So, I am actually observing opportunities for firms like SG Analytics here, where companies who have a certain use case or a problem statement are coming and discussing that with us. And once we incubate that conversation within our framework and see the success of it before they want to bring it back to the system and scale it up.

So that’s one immediate opportunity as well as a phenomenon that I’m seeing in the market.

AIM: How does SG Analytics ensure the privacy of data when dealing with the extensive data requirements of Gen AI, which relies heavily on large volumes of data?

Sid Banerjee: It’s good you’re asking that question to a company that is very passionate about ESG, encompassing both the social and community aspects. This is something that worries us as well. So let me approach it in three aspects. One is the impact of this on intellectual property rights. The second could be in terms of the impact on the organizational standpoint, and third, on the social and community standpoint.

Let’s look at deploying this in an area like the European Union, where GDPR is such a big concern. The European Union has signed up for GDPR, stating that no personal data should come out of Europe, if at all. So, you need to anonymize the data, mask the data, bring it out, and again, de-anonymize and bring it back. Given that sort of scenario, it becomes very restrictive for that market. And we have to be very careful if we have to deploy any of those. As a result of which, today, we have started an office in Wroclaw in Poland because we do not want the European data to come out of that.

On the intellectual property side, when Gen AI gets into managing this data and gaining insights, one has to be very careful in terms of potential patent or copyright violations in the process. Sometimes violations get morphed into the output without immediate awareness, until someone points it out.

Equally, from an organizational impact perspective, there’s standard chatter that this will have a significant impact on certain job sectors, which is a little premature for us to draw any conclusions. But it is creating some sort of panic within the system, questioning whether certain jobs or organizations will become redundant. These conversations are premature. We haven’t seen the full breadth of its output.

The same applies to the social and community side. Some people think this will lead to more carbon emissions, or that it may produce information and insights that could harm certain communities or their prospects. I believe this perspective is quite misplaced. With increased awareness and the presence of more watchdogs and governing bodies, I think they will certainly implement the right restrictions and regulations to prevent any misplacement.

AIM:How do you envision the ethical and fair use of Gen AI in the broader context of enterprise adoption, and which stakeholders should be involved in shaping its implementation?

Sid Banerjee: This reminds me of the old adage “That fire is a good servant, but a bad master.” Gen AI kind of fits into the same place. Left to wrong hands and left for the wrong purpose, it can have a disastrous effect. So I definitely believe that it is important to understand the right use case for this and use it in a manner which will make life possible as we talked about in SG. It will improve life, it will reduce the gap between the privileged and the underprivileged and it will make life much more beautiful if I may use the word. As against the concern about the whole disruption that will create something very similar to what we saw when initially automation came into play in late 80s and early 90s.

So overall, I think Gen AI is going to be a game changer. But it would need to be driven from the top. And that’s one of the reasons why from an SGA standpoint, I relocated myself as CEO of the company into the US as you are seeing that I am right now sitting in New York. And I want to engage with the CXOs of large transnational companies and work with them and talk through and drive these issues with much more impact and effectiveness and really make a difference in the way people will look at Gen AI in the future.