The shipping giant, which already deploys artificial intelligence in software development and other areas, is now looking to drive AI agents further into operations, including network planning and business processes. By 2028, FedEx expects to have AI integrated into more than half of its core operational workflows. FedEx is currently focused on setting up the underlying data and management foundation to oversee its AI bots.
Though logistics providers like FedEx are aiming to adopt AI, they’re grappling with challenges like managing numerous, disconnected data sources. “Logistics can be very fragmented—especially if you think of a global organization with their network being everywhere, it makes it difficult to standardize,” said an industry consultant.
As its underlying tech is completed, FedEx expects to roll out AI and AI agents that connect macro and microeconomic trends to better plan its network. In marketing and campaign management, FedEx will create a hierarchy in which there’s a “manager agent,” an “audit agent” and a “worker agent.” The goal of the hierarchy is to ensure that the agents have a trail of accountability for their actions.
At the moment, FedEx’s enterprise data platform, called Atlas, supports more than 200 AI use cases across the supply chain, commercial teams and enterprise functions. It has already turned on AI agents in areas such as software development, where they are developing and testing code. In operations, agents are helping customers clear customs more quickly.
Plans for FedEx’s AI agents also involve getting its humans ready to interact with the technology. the company just launched an AI education program for 300,000 of its employees, as well as a more advanced version for its technology workers. Each employee received a customized training depending on their role. FedEx says it doesn’t plan for those agents to replace its workers.
Classroom discussion questions:
- Why is FedEx pushing for more AI agents?
- How will agents be used in operations?
In our latest podcast, Barry Render interviews John Dyer, a well‑known speaker, consultant, and expert in continuous improvement, and the author of 
The global map of robotics is specialized. There is a multi-polar supply chain that is difficult to disrupt:
This forces a shift from one supply chain to a portfolio of capabilities designed around distinct value propositions including speed, reliability, customization, cost-to-serve, and compliance. Where commercial commitments are made in isolation from operations, the consequences surface later through margin erosion, excess inventory, and lost customers.
For two decades, manufacturing has been defined by a relentless pursuit of optimization. We automated assembly lines (Ch. 9), digitized records and built predictive maintenance models (Ch. 17), all in the service of marginal gains in efficiency.
By modeling factories and distribution centers digitally before making physical changes, PepsiCo hopes to cut down on costly mistakes while improving speed and capacity.
Manufacturing faces a dual disruption. AI, robotics and automation are reshaping production at unprecedented speed, while skilled labor shortages intensify when experienced workers retire, taking decades of knowledge with them. 
Scale drives efficiency—for almost a century, industrial planners have relied on this simple principle. In 1936 aeronautical engineer Theodore Wright discovered that costs fell in a predictable way every time production doubled. The more you produce, the cheaper things become, in part because the learning cost per unit declines. This is the topic of Module E in your text.
Several researchers compared the circumstances to war. “We’re basically trying to speedrun 20 years of scientific progress in two years,” said one Anthropic scientist. “Extraordinary advances in AI systems are happening every few months. It’s the most interesting scientific question in the world right now.”
