McKinsey & Company’s seminal study titled “Disruptive technologies: Advances that will transform life, business, and the global economy” listed 12 technologies that could “drive truly massive economic transformations and disruptions in the coming years”.  Second from the top, with an estimated economic impact of between $5 trillion and $7 trillion annually by 2025, was the automation of knowledge work.

“Knowledge work” involves complex analyses, subtle judgments, and creative problem solving, and encompasses a wide range of professions that include customs brokerage.

The prototypical knowledge work performed by Customs Brokers is Harmonized System (HS) commodity classification. Proper and efficient HS classification requires profound skill and years of experience. Still, studies have revealed disturbingly high rates of error – several customs authorities report error rates of 30% and higher – which can be attributed to both the complex nature of the task and the fact that it remains a mostly manual process.

Until now, the main challenges to cost-effective and accurate HS classification have been: 1) proper consideration and application of the HS General Interpretive Rules and Legal Notes; and 2) making the link between how products are described commercially by traders (e.g. “baby food”) and how they are expressed in the HS nomenclature (e.g. “homogenized composite food preparations”).

These challenges have largely been addressed with the advent of purpose-built tools that incorporate AI and machine learning techniques.

In “FuturaCorp: Artificial Intelligence & the Freedom to be Human”, IPSoft describes a future workplace where humans and machines work in concert to increase output. The report defines three tasks requiring a different mix of human and artificial intelligence, and predicts that 80% of deterministic tasks (i.e. repetitive and process oriented) will be done by machines in the not-too-distant future; probabilistic jobs (i.e. jobs that require humans and machines to interact) will be shared 50:50; and humans will perform 80% of cross-functional reasoning tasks (i.e. task that require connections that can only be made by the human brain).

The IPSoft model reveals how the Customs brokerage industry, and particularly the task of assigning HS codes, might be transformed. Since the vast majority of goods can be classified deterministically, 80% of classifications could be done by either machine alone, or by humans interacting with machines.

Even as AI & Machine Learning become ubiquitous among the knowledge-based professions… and they will… there is still no real expectation that the professional customs broker or clearing agent will disappear. What is certain is that smart technologies will eliminate the need for experienced and expensive staff to perform routine tasks, freeing them to work on problems that truly require their expertise and add further value to their organizations.