AI customs tools blamed for errors as calls for rules rise
AI tools are struggling to keep pace with rapidly changing customs rules and complex cross-border requirements, according to trade compliance specialists who report a rise in businesses seeking help after automated systems have led to delays and penalties.
Importers of Record (IOR) and Exporters of Record (EOR) at Mouse & Bear say they are seeing more companies whose compliance standards have slipped after adopting AI-based tools for customs and trade processes.
At the same time, senior data executives argue that regulatory frameworks will be essential to build trust in AI more broadly, as policymakers, including MPs in the UK, push for tighter oversight of the technology.
AI in customs
Mouse & Bear, which acts as IOR and EOR for technology firms shipping hardware and equipment globally, says some businesses have relied too heavily on automation to manage customs requirements.
"Businesses assume that AI is ready for every challenge. However, what many have experienced is that AI is slow to keep up with changing regulations and lacks the human element needed to navigate complex trade agreements. This has resulted in delays and fines on shipments," said Gemma Stunt, Sales Manager, Mouse & Bear.
Research from VAO indicates that 23% of global customs operators were using AI in 2024, and that this figure is expected to reach 45% in 2025. Many logistics and trade teams have introduced AI in an effort to accelerate document processing, classification and routing decisions.
Mouse & Bear argues that customs rules often contain grey areas that are difficult for automated systems to interpret without human oversight. Regulations differ significantly between jurisdictions and are updated frequently, which can make it hard for models trained on historical data or static rule-sets to produce accurate outputs.
Complex requirements
Stunt said that the consequences of errors can be significant if shipments are held at borders or misdeclared.
"While AI is a useful tool for many business operations, it may not always be the most accurate when it comes to more complex requirements and, in customs compliance, accuracy is everything. Mistakes in the customs process can lead to shipments being delayed or seized with costly consequences. That's why having an experienced team to navigate these complexities is critical," said Stunt.
Mouse & Bear supports technology companies moving items such as computers, IT hardware, telecoms equipment and medical technology into more than 160 countries. The firm provides tailored import plans and customs support, aiming to maintain visibility and control over clearance processes.
Stunt said that demand for this type of assistance has grown as organisations recognise the limits of automation in areas that require judgement and up-to-date regulatory knowledge.
"As specialists in IT hardware and technical equipment, we've seen firsthand how technology has become more powerful in recent years. However, in customs operations, it can only support - not replace - real, human expertise. That's why we're seeing more businesses turn to us to help them navigate the customs processes, as it's our human experience that ensures their shipments arrive without complications," said Stunt.
Regulation and trust
While some trade specialists stress the need for human oversight in compliance, others in the data and analytics sector say that appropriate regulation will be key to expanding AI adoption responsibly across industries.
Jane Smith, Field Chief Data & AI Officer at ThoughtSpot, said that previous waves of regulation in sectors such as banking and aviation had supported growth by providing common standards and reassurances for users.
"History shows us that regulation doesn't kill innovation; it enables it. Banking couldn't have reached today's scale without deposit guarantees that gave people confidence to trust their money with financial institutions. Aviation became a mass market industry because of safety standards that made millions comfortable flying. Rather than being seen as an obstacle, regulation provides the legitimacy and trust needed for widespread adoption of AI, ultimately making the technology more valuable.
Today, AI represents the biggest catalyst for growth, but there's a critical barrier - trust. The black box approach of many AI systems creates hesitation. Smart regulation creates the transparency and explainability frameworks that unlock AI's true potential. When regulation establishes clear rules around data access, portability, and AI explainability, it enables the secure sharing and analysis of data that AI needs to thrive. Instead of fearing opaque systems, organisations gain the regulatory legitimacy to adopt AI responsibly at scale," said Smith.