DSCInsights in Action: AI Regulations Impact on Global Supply Chains

Welcome to the next episode of Digital Supply Chain Institute insights in the in action. In today's episode, we are talking about AI regulation. Now, not talking just through a perspective of high-level ideas or you know what's going on, we have a tangible thing which has been produced by Digital Supply Chain Institute and it's an article which really points out the AI regulation framework and on the other side how it's cut across supply chain. So, my guests today are Sarah Latif and Darinda Shataram who are members of DSCI team and both of them co-authors of the article. And what we wanted to do is drive you through the key learnings cutting across various different industries and continents related to AI regulation. So, DS, Sarah, welcome. >> Thank you. Thank you. >> Excited to be here. >> Let's start first with you, Sarah, and let's, you know, build on what I gave as an introduction. You know, the article points out that the AI regulation is no longer just a legal issue, right? It's a core operational issue for supply chain. So, from what you have learned, uh what are the the biggest mistakes organizations are making and uh what they still treat like compliance as a Do they still treat compliance as a reactive function instead of the strategic business capability? How they balance these things? >> Yeah, thank you, Marco. So, I think there's two big mistakes that companies are making when treating compliance as an afterthought rather than designing it into the solution at the beginning. So, AI regulations now impact all aspects of how a supply chain operates from day to day. So, they influence what data can be used, where it can be stored, how it's processed, what level of human insight is required, and even how supply chains work with their third parties, and you know, such as their suppliers and partners. So, waiting until deployment to consider these requirements often times will result in companies spending more time and more resources redesigning systems, um requires them to delay rollouts, and often times even facing for non-compliance. So, those that factor in compliance from the start, they'll save money, they'll save time, uh they'll save resources, and even create a competitive advantage for themselves compared to those that have to then backtrack and try to adjust to become compliant. And then another mistake um is reading compliance only as the responsibility of the legal department or team. As AI systems become more advanced, they impact a variety of different functions within one supply chain. Um and the organizations that are the most successful are those that view it as a business capability. Um so, they'll build it into their operating model from day one, and then work together to stay up up to date on the latest laws and regulations. Um so, the strongest organizations really will shift their mindset from how do we comply to how do we build a competitive advantage while being compliant? >> Thank you, Sarah. And I think this brings us, you know, like every time we look into a broader context like you did here, it we always need to connect it with, you know, where it brings things to the to to the growth. And this leads me to the second question which I would like to pose to you, DS. You describe a growing competitive asymmetry between highly regulated regions and more flexible innovation-oriented regions, right? And we know that the world is divided in these two. So, how do you see this divide the supply chains and reshaping global competition over the next few years, especially in multinational companies. Because these things should be aligned and the growth should be built around it, but again, you have two different realities. >> Marco, thank you for asking the very important question. The focus is on where you operate and what can you do with the AI. As the paper described, there are two sets of frameworks. One very rigid, very highly regulated. Other side, we have the flexible and then innovative framework. So, it depends upon how the company operates in these both these environments. So, some of these things, some of the organizations are trying to write do pilots in, you know, permissive governments and permissive frameworks, actually. So, the challenge for these supply chain companies would be how they would be able to balance out, you know, how they operate in the highly regulated market with respect to the, you know, flexible market. So, this is the lesson that these are they would learn as you as they go start implementing their, you know, frame their solutions in these environments. >> Thank you. Yes, and I think building on on that again, like when when you need to connect things horizontally and having the same organization two different uh uh polarities, if we can put it that way. That that that leads into uh a thinking of the synergy and how it can how things can be modular. So, Sarah, question for you. One of the points as well in the article is actually a need for modularity, right? Adaptable technical architecture. So, what does that look like in practice? I think it it's a high-level conclusion, which is important, but for supply chain leader who is trying to scale AI across multiple countries with conflicting regulations. How do you proceed? >> Right. So, I think it's all about designing AI systems with flexibility in mind from the get-go. Historically, organizations tried to build one global technology solution and then implement it worldwide, which is now becoming increasingly difficult and inefficient as there's different regulations for different geographical regions. So, a modular architecture would then allow an organization to maintain a common AI foundation while having those kind of adaptable features that will allow them to apply their to meet compliance that ranges in different geographies. So, for example, a company would have, let's say, a certain forecasting engine globally, but apply different governance structures to it depending upon where they're actually operating. And so, from a leadership perspective, it's really about building adaptability into an organization's technology strategy, so they don't have to redesign, you know, from the ground up depending upon the regulations of the market they're operating in. >> And the combining all the all the three questions, we always like to wrap up with the practical steps. And I think, you know, it's always great to bring the learnings to say which direction the practical application can go, but on the other side, it always resonates with people with, you know, what's the first step I can do. So, DS, a question for you to wrap wrap up the conversation is, you know, you emphasized that balancing rapid AI innovation with governance and risk control, which are two important pillars for having successful AI strategy, what right? What are the practical first steps supply chain executives should take today to prepare their organization for the future of global AI regulation? >> Again, thank you for asking me to sum this up, Marco. There is no a space for something called regulatory certainty. You know, regulations are evolving. If you followed the news last couple of weeks, you know, US government has released a new set of regulatory guidelines on as far as the AI is concerned. So, as this regulatory structures are evolving around the globe, three things need to consider for a supply from from point of view from supply chain point of view. One, governance need to be a cross-functional. That is, technology cannot be the only pillar of in this case. Legal, procurement, risk, you know, operations, all the departments in an organization need to get involved in this understanding this AI governance or AI regulation. Number two, you need to build foundational capabilities, you know, things like documentation, traceability, monitoring systems, you know, and the typical what you have for any certification process you have to you need to build in the organization. Third, of course, is the talent, you know, people should understand what this governance means, what this regulation means to her or his job as I think. So, it's a very important to build this sort of a culture in the organization to understand what AI regulation means, how it affects him or him at the at the job. Yeah. So, the goal is the organization can innovate, you know, scale responsibility, and adjust as a regulatory environment proceeds. Like I said, you know, there is always a challenge between a oversight and innovation. That's the struggle that has to go through for a long time to come. Thank you for the asking this question, Marco. >> Thank you very much, DS, and thank you, Sarah. I think this It a very productive conversation. On the other side, I do know that uh this work will be continuing because uh we can only say that uh the AI regulatory side is super important, but it's still evolving. And it's evolving alongside the innovation. And I think what company what what you have shared with us has been very valuable because the key thing is that we can say that this is an ongoing process. It won't be finished in next few months. It will be definitely building up over the years as far as technology accelerates. And on the other side, as far as we find more applications which are valuable for supply chain transformation. So, the key notions are keep on experimenting. Keep on following where the regulations are in different capacities and areas and geographies you are present. And uh do not be afraid to innovate. And with this, I think we can uh close this conversation and say uh I would like to thank two of you again for joining us. I would like to thank our followers for uh being with us. And we keep the promise that once you come up with the next phase uh let's say uh engagement and learnings from the AI regulatory field in next 6 to 8 months, we'll definitely reconvene back and uh build up to our followers where the things are going. So, once again, thank you very much. And this was another episode of Digital Supply Chain Institute Insights in Action. And we'll come back to you with more interesting topics to assess. Yeah.