ASYCUDA founder warns legacy Customs systems obsolete in LLM era
Jean Gurunlian, Chairperson of Webb Fontaine and the architect behind the ASYCUDA Customs system, issued a stark warning on the future of Customs technology at the 2026 WCO Technology Conference in Abu Dhabi. He stated that legacy Customs platforms are no longer fit for purpose in a trade environment reshaped by large language models (LLMs).
Gurunlian, whose ASYCUDA system has been deployed in over 100 countries, said traditional Customs systems are constrained by static rules and slow update cycles. These platforms are increasingly incapable of keeping up with fast-changing regulations and complex trade policies.
He emphasized that adaptability is no longer optional. “Customs systems that cannot adjust to new laws or operational requirements within very short timeframes simply will not survive,” he said, highlighting the structural limitations of rule-based architectures.
LLMs expose weaknesses of static platforms
According to Gurunlian, the rise of LLMs has revealed the fragility of systems built on predefined logic. Many platforms still take months or even years to integrate legislative updates, tariff adjustments, or new non-tariff measures, leaving administrations vulnerable to sudden policy shifts.
“Tariffs and non-tariff barriers can change overnight and are increasingly used as political tools,” Gurunlian explained. “LLM-enabled systems can interpret and operationalize these changes in seconds, which traditional platforms simply cannot achieve.”
He stressed that modern Customs technology must be built around continuous learning, contextual awareness, and rapid adaptation—capabilities that only AI-driven architectures can provide. Systems that cannot improve in production, he argued, should not be deployed at all.
Governments face a critical technology decision
Gurunlian warned that continued investment in static, rule-based platforms risks locking governments into technology unable to respond to geopolitical shocks or regulatory volatility. He noted that the future of trade depends on systems that can evolve alongside global complexity.
“LLMs change the very nature of Customs systems,” Gurunlian concluded. “This is not about adding AI to existing platforms. It is about designing Customs architectures from the ground up to be adaptive, intelligent, and resilient.”
