Anyone even remotely plugged into the news would understand at some level that artificial intelligence (AI) is poised to unleash massive economic change. The idea that container shipping and supply chains will somehow be unaffected is fanciful.

Evidence of the huge potential impact is everywhere.

The CEO of Amazon said in June the company will need fewer employees in the future — certainly without any curtailment of revenue growth — while a month later, the CEO of Ford said AI will replace “literally half of all white-collar workers.” The stock price of the tech research firm Gartner collapsed 30% on Aug. 5, its worst single-day loss in 26 years, after earnings weakness pointed to the vulnerability of its business model to AI. The share price of the advertising firm WPP was down nearly 50% this year as of Sept. 4 on fears creative services will be displaced by AI. How many coders will be needed if the work can be done by AI at a fraction of the time and cost?

The impact on container shipping could be similarly profound, but not in obvious ways.

Of course, there will be an impact from human processes being automated with AI. That is a given. Carriers and forwarders know they can’t miss any opportunity for cost savings, not just to eke out profits, but to avoid going to market with a higher cost structure than their competitors. This is nothing new. AI will be just the latest technology leveraged in the pursuit of cost competitiveness.

No, the impact could be much greater.

Ocean carriers increasingly resembling railroads

The late E. Hunter Harrison, CEO of Class I railroads Canadian National, Canadian Pacific and CSX, was famously beloved by investors and loathed by customers. Running his railroads to maximize profit meant customers, specifically those captive to railroads, were inconvenienced and deprioritized in ways large and small. Precision schedule railroading, a concept he created that has been widely implemented, meant trains left on time with no exceptions, even for preferred customers.

Ocean carriers increasingly resemble railroads. Beneficial cargo owners are beholden to container shipping; only for a few and only in certain situations is air cargo a viable alternative. And with waves of consolidation over decades, top carriers control more of the market than they ever did. The top 10 now control over 84% of total tonnage, according to Alphaliner.

Although carriers freely compete, the COVID-19 pandemic changed their psychology. This is relevant in light of AI.

Consider the evolution.

In an earlier day, with more carriers and an ethos of serving customers, carriers used to run their networks for schedule reliability. This was standard during many decades of the conference system, when conference lines spent lavishly on fuel to ensure on-time arrivals regardless of whether freight rate levels would yield adequate returns.

This is no longer the case. Devotion to reliability was punctured during the 2008–09 financial crisis when carriers began slow steaming at scale to balance supply and demand. It was further eroded in the years prior to the COVID-19 pandemic as blank sailings became standard practice within alliances.

The pandemic and the disruptions that followed sealed the deal. Carriers realized that profits result not from ships being on time with ships and cargo circulating fluidly. Rather, profits result from the system being disrupted and capacity removed due to port congestion, idled ships and elongated transits, like those around southern Africa to avoid the Red Sea. On that basis, carriers earned unimaginable profits of over $400 billion from 2020 to 2023, according to analyst John McCown. That was more profit than the industry earned during the entire period between the first container ship sailing in 1956 and 2020, according to Sea-Intelligence.

AI to the rescue?

If JPMorgan is right in recently forecasting “deep losses” due to “material oversupply,” carriers are staring at a bleak outlook over the next few years. Could this be where AI comes to the rescue? Its timing couldn’t be better.

The following scenario is far from implausible.

A carrier’s complete data repository is fed into a trained AI model on a continuous basis: all vessels, all locations, all speeds, all fuel consumption, all containers and their locations, all customers, all contracts, all spot rates, all port times and all financials.

A simple request is made: Run the network to maximize profit. Do as Hunter Harrison would have done. Determine at what speeds ships operate, the ports they do and don’t call, and which trade lanes they slot in and out of. Also determine which customers get containers and when, and which ones must wait, and what rates get extended to whom, when and on what terms. The system learns and improves.

Google says its Shipping Network Design API, announced last year, is able to double the profit of a container line, deliver 13% more containers and do so with 15% fewer vessels. In April, CMA CGM announced a €100 million investment in a partnership with Mistral AI that CMA CGM Chairman Rodolphe Saade called a “transformation of CMA CGM through artificial intelligence.”

In a famously relationship-based industry, subjecting key decisions to AI direction is obviously radical. Carriers may hesitate to subjugate customer relationships to AI decision-making if competitors are keeping the phone lines and bar tabs open. They may defer to their conservative instincts, just as they do regarding other proposed radical solutions to underperformance, such as trading freight rate futures.

But times have changed. Carriers are no longer unclear that they are profit-making entities entitled to act in their own interests. Transportation profits come from control over the network, limited competition, capacity management and captive customers.

That is what is developing in containers, and at the end of the day, AI will only accelerate it.