Latest / Global Business insights / Artificial Intelligence and War Reshaping Supply Chains

Transcript

Artificial Intelligence and War Reshaping Supply Chains · Global Business insights

• 0:06
  SPEAKER_00: Artificial intelligence and changing geopolitics are driving a major shift in global supply chains.
• 0:16
  SPEAKER_00: Companies are moving away from traditional cost-focused approaches and building more resilient, autonomous and regionally diverse networks.
• 0:28
  SPEAKER_00: Analysts say organizations are speeding up AI adoption to handle complicated tasks, making technology and political strategy essential parts of logistics.
• 0:40
  SPEAKER_00: The era of procurement as a backroom rule is over.
• 0:44
  SPEAKER_00: Now it's a strategic lever for transforming performance with sea level visibility.
• 0:50
  SPEAKER_00: Getting it wrong can sink your business.
• 0:53
  SPEAKER_00: Ignoring shifts caused by challenges like we saw in COVID, workforce aging, inflation tariffs, and geopolitical changes will leave you behind as others use AI to build resilient supply chains and make progress.
• 1:10
  SPEAKER_00: Recent PMI data, especially for manufacturing, shows that geopolitical uncertainty is prompting companies worldwide to accelerate procurement in at its anticipation of price hikes and supply shortages, temporarily boosting economic activity.
• 1:28
  SPEAKER_00: In the US last month, purchasing growth reached a four-year high, driven by supply delays and inflation pressures linked to the Middle East conflict.
• 1:39
  SPEAKER_00: Businesses are not just buying more, but leveraging generative AI for smarter, more efficient procurement as other costs rise.
• 1:51
  SPEAKER_00: Hello, this is Global Business Insights, where we bring perspectives that help to navigate the more challenging context.
• 1:59
  SPEAKER_00: This episode focuses on reimagination of procurement and supply chain.
• 2:04
  SPEAKER_00: Tosen Ifashide is a procurement and supply chain expert specializing in strategic sourcing, supplier negotiations, and cost optimizations across complex global environments.
• 2:19
  SPEAKER_00: He comments on how AI is rewriting procurement and supply chains.
• 2:25
  SPEAKER_01: Supply chain has gone through a journey, starting from you know the old vertical integration, asset ownership.
• 2:33
  SPEAKER_01: We have the lane just in time.
• 2:35
  SPEAKER_01: We look at Kanban, waste reduction, link Six Sigma.
• 2:39
  SPEAKER_01: Then we've got the materials requirement planning to ERP, enterprise resource planning soft system, which is used for the computer generation of material planning.
• 2:50
  SPEAKER_01: Then we've got globalization, which looks at continentalization, global uh value chain, and as well as outsourcing.
• 3:00
  SPEAKER_01: Then we also have digital supply chain that looks at the cloud, internet of things, control towers.
• 3:08
  SPEAKER_01: And the last one but not the least is AI, which looks at machine learning, generative AI, digital twins, and agents, which we'll talk about more.
• 3:19
  SPEAKER_01: How does predictive intelligent work or AI?
• 3:24
  SPEAKER_01: There are two sources: internal data, external signals.
• 3:30
  SPEAKER_01: Internal data is what's on the ERP, inventory, uh planning, uh what we call uh um the the planning uh you know the long-term, mid midterm planning.
• 3:45
  SPEAKER_01: They also have external signs, which is you know what's going on the market, the news, the weather, and all the macroeconomics.
• 3:56
  SPEAKER_01: When they all mix together, they come, they did, they they constitute pattern recognition.
• 4:03
  SPEAKER_01: AI, and that's where AI is very good at, it looks at pattern recognition, and based on that pattern recognition, it comes up with a forecast, with a confidence within a range, and based on that forecast, you now can take action to replenish or escalate.
• 4:25
  SPEAKER_01: But here is the thing you still need human intervention, and that will come up later in this conversation.
• 4:32
  SPEAKER_01: Human intervention is required.
• 4:35
  SPEAKER_01: So there are several aspects of AI generation, uh, which are this five aspects.
• 4:44
  SPEAKER_01: First of all, this machine learning, and what does machine learning mean?
• 4:48
  SPEAKER_01: It predicts patterns, it detects signals from structured and historical data.
• 4:54
  SPEAKER_01: One example I'd like to give is requirements planning.
• 4:59
  SPEAKER_01: It looks at consumption history.
• 5:01
  SPEAKER_01: If you have uh your long-range planning and everything, it looks at that, it looks at different areas and it can determine the mean max level, the reorder planning, and completely rewrite the stocking strategies so that it's not using stale information, but it will look at plan delivery time, it looks at what's happening, it looks at vendor behaviors, it looks at all these other things, as well as purchase patterns, price determination, and maps a way out.
• 5:32
  SPEAKER_01: So, what it does is it improves stock ass forecasting, it reduces stockouts, excessive venture, strengthens suppliers, risk, resilience, and helps in spend visibility.
• 5:47
  SPEAKER_01: The second one is natural language processing.
• 5:50
  SPEAKER_01: This one interprets, reads data, and information, and it's good from a uh clause clarification, looks at contracts, what is able to highlight the risk areas, um, it it interprets the PO text, it also can categorize spend based on text information and can determine what the category is, and it can track contract creation and renewal.
• 6:19
  SPEAKER_01: That's what natural language processing, and it's also helpful in redlining contracts.
• 6:26
  SPEAKER_01: We have generative AI, which is the content creation, it looks at procurement intake, it looks, it summarizes a draft, it explains, it compares, it converts unstructured data to structured.
• 6:40
  SPEAKER_01: Those are some of the areas where you can even and write some stuff and it still gives you a summary on it.
• 6:46
  SPEAKER_01: So this one reduces reading time and drafting time.
• 6:49
  SPEAKER_01: You can quickly summarize.
• 6:51
  SPEAKER_01: Some people have taken videos and loaded it, and it just gives them a summary of all the discussion.
• 6:56
  SPEAKER_01: It also generates meeting meeting minutes if you have a transcribed voice, and it helps in faster decision making.
• 7:06
  SPEAKER_01: The fourth is digital twins, this which we call simulation, it takes it up to a whole level where you simulate different scenarios, whether mostly in logistics, and you look at and then you now do what ifs and all those things in there and can enable you to simulate the optimal process rather than do try and error.
• 7:30
  SPEAKER_01: Then the last one is robotic process automation, which is RPA, which looks at able to be able to autonomously go through certain transactions.
• 7:41
  SPEAKER_01: This still at its infancy and has to be handled um carefully, but it's there and it can reduce manual effort cost processing errors and improve cycle time.
• 7:55
  SPEAKER_01: What are the improvements that we've seen?
• 7:59
  SPEAKER_01: First of all, forecast error reduction is 35%, inventory um improvement is by 25%, then we also have logistics cost reduction around 10-15%.
• 8:15
  SPEAKER_01: Uh, we've got um admin reduction, work reduction at 33%.
• 8:21
  SPEAKER_01: Now, the king of everything is a contract review, a whooping 70%, and there is an easy explanation for that because over time a lot of organizations have been successful in using co-pilots and everything to review red line contracts, give it the red line, and legal would do the finishing touches with it, which has reduced the administrative uh time on contract review.
• 8:53
  SPEAKER_01: So it's really a very big success story.
• 8:59
  SPEAKER_01: There are success stories across credible companies like Google, Amazon, Wells, Fargo, and Microsoft.
• 9:08
  SPEAKER_01: They've all had contract review reduction, um, inventor reduction, better forecasting accuracy, and hours saved manual in procure-to-pay process.
• 9:19
  SPEAKER_01: These are publicly um available time that has been able to prove that we have some success there.
• 9:29
  SPEAKER_01: It would be unfair to talk about all this beauty of AI without talking about inherent risks.
• 9:37
  SPEAKER_01: And so some of the areas is I'll talk about data.
• 9:44
  SPEAKER_01: Data, wrong data will cause hallucinations.
• 9:48
  SPEAKER_01: Um, for example, you write a free text and you don't put the right information there.
• 9:55
  SPEAKER_01: AI cannot tell you what the correct categories or classification.
• 10:00
  SPEAKER_01: AI does not read people's minds, neither can a human being do.
• 10:04
  SPEAKER_01: It's what you write down that you will interpret.
• 10:07
  SPEAKER_01: So that's another thing.
• 10:09
  SPEAKER_01: Then also don't put it's suboptimal to put AI in front of a non-transform process.
• 10:18
  SPEAKER_01: So you need to transform your person and um streamline it for best results.
• 10:25
  SPEAKER_01: There's privacy and security issues, and there's also um user um phobia on it as well.
• 10:35
  SPEAKER_01: Now, what are the three ways of doing it?
• 10:37
  SPEAKER_01: Because for user prophobia, one of the things is people there's word out there that AI reduces jobs, so there's a resentment of AI, and at times when you bring it in, people look at it from s with suspicion, and there may be malicious compliance or even sabotage in some cases, in extreme cases, anyway.
• 10:58
  SPEAKER_01: So, what do you do to correct that?
• 11:00
  SPEAKER_01: First of all, you look at the policy.
• 11:03
  SPEAKER_01: What do you use AI for and what you don't use AI for?
• 11:08
  SPEAKER_01: Secondly, there's still human intervention to approve certain things.
• 11:12
  SPEAKER_01: AI is something you can teach, show the right way to do things that you perfect it.
• 11:17
  SPEAKER_01: So you still need human intervention to look at things, especially on high profile areas.
• 11:22
  SPEAKER_01: Then you also have data controls access, you ring fence certain things.
• 11:28
  SPEAKER_01: Uh Microsoft Core Pilot is very, very good at that point where you can ring fence certain areas that you say it does not the data doesn't leak outside as well.
• 11:39
  SPEAKER_01: So here's a summary within this.
• 11:42
  SPEAKER_01: A uh governance is not a break in AI value, it's actually a strategy that makes it sustainable because you start, use governance to control, make it scalable as it's successful in small things, then it becomes unstoppable as you keep on using it.
• 12:01
  SPEAKER_01: Rather than go big and then you flame out and you go back to pre-AI time because an incident occurred.
• 12:09
  SPEAKER_00: To other stories, Anthropic is partnering with major private equity firms, including Blackstone, Goldman Sachs, and Hellman and Freidman in a$1.5 billion joint venture to help businesses adopt its AI tools.
• 12:25
  SPEAKER_00: OpenAI is also pursuing similar deals with Wall Street.
• 12:31
  SPEAKER_00: The latest April PMI data is giving the first clear reading on the initial impact of the war in Iran and other developments on economic activity all over the world.
• 12:42
  SPEAKER_00: Euro area economic growth was weak in Q1, with GDP rising just 0.1% quarter and quarter.
• 12:48
  SPEAKER_00: April's composite PMI, that's Europe, dropped to 48.6, pooled down by services PMI at 47.4, while manufacturing improved to 52.2 due to stockpiling and healthier backlogs.
• 13:04
  SPEAKER_00: Growth expectations remain subdued.
• 13:07
  SPEAKER_00: US manufacturing expanded for the fourth month with ISA manufacturing PMI at 52.7 and SP global PMI rising to 54.5, the highest since May 2022.
• 13:20
  SPEAKER_00: The SP Global US services PMI was 51.0 in April, up from 49.8 in March.
• 13:29
  SPEAKER_00: The SP Global UK manufacturing PMI rose to 53.7 in April 2026 from 51.0 in March, exceeding expectations.
• 13:41
  SPEAKER_00: Output expanded for the sixth time in seven months due to rising domestic and export orders.
• 13:48
  SPEAKER_00: In Africa, the economies are also showing strong resilience.
• 13:51
  SPEAKER_00: Nigeria, April PMI was up to 52.4 versus 51.9, though noting high energy costs are beginning to weigh on growth.
• 14:02
  SPEAKER_00: For South Africa, manufacturing PMI also went up to 52.6, seeing its first expansion in seven months.
• 14:10
  SPEAKER_00: Global stock markets remain volatile, Asian stocks reached record highs amid optimism for a resolution in Iran and strong demand for chips, while investors capitalized on dips following robust earnings.
• 14:26
  SPEAKER_00: AI continues to boost tech shares.
• 14:29
  SPEAKER_00: Brent oil stays above$100 per barrel, though crude is selling as traders expect reduced tensions in the Strait of Ormuz.
• 14:40
  SPEAKER_00: Trump halted US military guidance for commercial ships, citing progress in talks with Iran.
• 14:47
  SPEAKER_00: Well, that's it for this special edition of Global Business Insights.