Artificial intelligence (AI) is beginning to become a more practical tool in everyday customs and trade processes, but adoption among businesses remains cautious and uneven.
That was the picture painted by trade digitalisation experts on the Chartered Institute of Export & International Trade’s latest Global Trade Live webinar, where panellists agreed that the technology can reduce trade friction, improve compliance and help businesses navigate increasingly complex border processes.
They said that trade professionals are likely to go through an AI transition rather than transformation, with many organisations still testing tools and learning their limits, rather than embedding AI across day-to-day trade operations.
Compliance requires human oversight
Businesses also remain liable for customs compliance, making trade professionals necessarily cautious about what constitutes appropriate use.
Ilona Kawka, Imports Advisory Practice lead at the Chartered Institute, said that AI remained a “high risk venture”.
“Errors can have real-life consequences,” she said. “We need to remember humans have accountability because HMRC will not accept it if we say AI told us to do something.”
“We need to ensure we are doing all the responsible reviewing and approving, as well as understanding the end-to-end process. We need to remember that AI is there to assist us, but it cannot make its own decisions.
“So, we must tell AI how to think to reflect our process and compliance, and we also need to have total control of the process.”
She added that businesses need to be able to show their own reasoning for any AI-generated outputs, citing examples such as customs classifications or valuations.
She was joined on the panel by Ronald Langbauer, a senior sales manager at MIC, Jose Puyana, director of trade at Transform, and Dr Awinder Kaur, associate professor at the University of Warwick.
What the audience said
The greatest indicator of AI’s impact being more transitional than transformation came from the audience itself.
About half of the audience (49%) said they were already using AI in customs and trade processes to some extent, but only 6% said were using it ‘extensively’. Just over a quarter (28%) said they not yet using it but were planning to, with 6% saying they had no plans to use it at all.
Delegates typing in the ‘chat’ area of the webinar said that they were already using or considering using AI to support reviewing documents, classifying products for customs and monitoring regulatory changes.
The audience was broadly positive about the possible impact of AI on their trade, with 52% saying there were ‘quite optimistic’ and 9% ‘very optimistic’. The remainder said they were ‘not sure’, with no delegates saying they were pessimistic.
Data concerns
Delegates said that ‘inaccurate data outputs’ was their biggest concern around using AI for customs and trade processes, with 29% saying ‘data reliability’ was their main issue.
Puyana said he wasn’t surprised by these concerns but said that businesses needed to “start thinking more around end-to-end process” rather than considering using AI for one-off tasks.
Using AI to support “the full decision journey will give help some of those concerns,” he said.
Skills and knowledge
One of the clearest messages from the webinar was that successful AI adoption depends less on the novelty of the tools themselves and more on the quality of the foundations underneath them.
Kaur suggested that the organisations that will benefit most from AI will not be those with the newest technologies, but those with “good-quality data, people who know how to use AI critically and leaders who are willing to keep learning” as both the technology and regulations evolve.
She added that appropriate use of AI could also allow trade professionals to more fully utilise their expertise on tasks where human oversight is needed.
“The real opportunity isn't just about really saving time. It's about giving trade professionals more capacity to make more complex decisions where human expertise, judgment and experience make such a big difference.”
Business usage
Langbauer added that “organisational setup”, including policy and governance, was needed to ensure trade teams are using AI in an effective way.
“There's a discovery phase that may be needed – a company needs to be set up to support AI properly, in ways of working that might be a bit different,” he said.
Langbauer said that a lot of the conversations he has with businesses are usually about identifying what problem or solution AI can tackle, rather than whether AI could be generally useful.
“There is a variety of fronts where it can really help,” he said, adding that those not currently looking at the technology “should have a look at adopting AI in the future, because it will certainly not go away”.
More to come
So where does that leave the sector today?
The state of play is best described as promising but still developing. AI is already helping some customs and trade teams work more efficiently, and there is broad recognition that it could play a much larger role in the years ahead.
But the panellists were clear that meaningful adoption will depend on more than enthusiasm. It will require investment in data quality, common standards, workforce capability and responsible governance.
AI in customs and trade is no longer a distant prospect, but neither is it a finished story. Its long-term value will depend on how thoughtfully and intentionally businesses choose to adopt it.
This article was written with the support of MS Copilot, an AI tool, but with editing and input from William Barns-Graham, head of content at the Chartered Institute and the host of Global Trade Live.
The next Global Trade Live webinar is on navigating disruption from this year’s crisis in the Middle East and is taking place on 14 July.
Member exclusive
