Artificial intelligence (AI) is quietly reshaping the logistics industry, with leading companies integrating advanced solutions in areas such as capacity and revenue management. However, according to Dilip Bhattacharjee, partner at McKinsey & Company, the real challenge is not the technology itself but the ability to drive effective change.“The larger, more established players in logistics have been quietly implementing AI in their functional areas,” says Bhattacharjee. “The issues remain the same – it’s not about technology but change management. It is the number one issue that we continue to hear about when we speak to customers.”While piloting AI-driven solutions is relatively straightforward, achieving impact at scale is far more complex. “Change management is the core enabler of impact – data is the backbone of effective AI solutions,” he said during a recent online event.According to Bhattacharjee, across the broader transportation sector, AI adoption has accelerated at an unprecedented pace. Generative AI (Gen AI), in particular, he said had gone from zero to thousands of use cases almost overnight. “Airlines, for example, have significantly improved disruption management and customer service by deploying AI to handle vast volumes of passenger interactions. Instead of thousands of human agents managing f light delays or rescheduling queries, AI-powered systems now synthesise and interpret customer needs. They can scale up from five calls to 5 000 in an hour. Humans simply cannot do that.”A recent McKinsey & Company study found that companies were increasingly using Gen AI in at least one business function. More than 40% of respondents indicated that advancements in Gen AI were prompting their organisations to increase overall AI investment. At the same time, the industrial robotics industry is projected to grow by more than 10% annually through 2030.Speaking specifically about the logistics industry, Bhattacharjee said that implementation remained a challenge due to the highly distributed nature of frontline operations. “The front line is so distributed, with not everyone in the same office, that the actual implementation of AI tools is not always easy.”He also highlighted a persistent issue in the sector: the difficulty of seamless data sharing across different stakeholders. “The notion of blind handoffs of information remains problematic, with one of the biggest challenges still being the sharing of data across different parties in the supply chain.”For smaller players, the hurdles are even greater. “Unless you are a large intermediary or a large carrier with significant amounts of your own data, it is just not that easy to implement AI,” he said. LV