Transform your trade compliance workflow: How AI eliminates the classification guesswork

Trade compliance runs on two related but distinct disciplines: assigning the right code to a product, and understanding what the regulations say once that code is assigned. Most AI trade compliance software on the market forces teams to stitch together separate point tools for each discipline. ONESOURCE Global Classification and Global Trade Research, both powered by CoCounsel, are purpose-built for each half of that work, designed to operate together inside the same ONESOURCE workflow.

Global Classification handles the classification decision itself: It ingests product descriptions and historical data, applies references to your organization’s own historical classification decisions to suggest HS codes, and surfaces recommendations for professional review, prioritized by how strongly the combined evidence supports each option. Global Trade Research handles what comes next: the regulatory research a classification decision often depends on, from tariff actions and executive orders to binding rulings and FTA eligibility, returning a cited answer instead of a list of documents to read.

Used together, the two tools close the loop between code and context: a classification decision backed by instant access to the regulatory reasoning behind it. That combination only holds up if it’s built on a foundation of trustworthy, verifiable content, starting with how ONESOURCE Global Classification puts that foundation to work today.

Jump to ↓
The hidden costs of a fragmented trade compliance workflow

How AI HS code classification software transforms the classification workflow

Real-world impact: From busywork to strategic value

The Thomson Reuters advantage: Trusted content meets classification technology

Why trustworthy AI starts with trustworthy sources

Your next step: Build a faster, more defensible trade compliance workflow

White paper   See how AI-powered classification compares to manual processes across speed, accuracy, and cost. Read the white paper ↗

The hidden costs of a fragmented trade compliance workflow

Time drain that steals strategic focus

Manual classification workflows can consume valuable time that organizations could redirect toward strategic priorities. According to Mariana Maldonado, Implementation Consultant at Thomson Reuters Professional Services: “Collaboration and knowledge-sharing across trade compliance and business teams remain fundamental to understanding and improving classification processes.”

Professional teams spend countless hours hunting through Federal Register notices, CBP messages, and tariff schedules, cross-referencing multiple sources for single classification decisions. Junior staff sometimes lack the expertise to handle complex classifications, while senior professionals can get bogged down with research requests that prevent them from focusing on higher-value trade optimization activities. And that research burden doesn’t end once a code is assigned: teams still need to confirm the regulatory reasoning behind it, often by digging back through the same fragmented sources all over again.

Risk amplification in a volatile regulatory environment

Misclassification penalties and supply chain delays create cascading business impacts. The HS system has 5,000 codes at the global six-digit level, expanding to over 19,000 unique codes for the U.S. schedule alone. With similar expansion across over 200 countries, maintaining accuracy while staying current with near-daily regulatory changes becomes an operational challenge.

Inconsistent classifications create audit vulnerabilities, while difficulty tracking regulatory changes increases compliance risk. Poor audit trails compound these challenges, making it harder to defend classification decisions when customs authorities make inquiries.

Operational inefficiencies are holding trade teams back

Decentralized classification processes lead to errors and compliance gaps across global operations. Data quality issues complicate accurate classification decisions, while fragmented workflows prevent teams from achieving the consistency required for multinational operations. The result? Trade professionals spend more time on administrative tasks and less time driving strategic value for their organizations.

How AI HS code classification software transforms the classification workflow

Technology that gets smarter over time

ONESOURCE Global Classification powered by CoCounsel targets 95% accuracy by learning continuously from your organization’s own historical classification decisions. When your team classifies a product, that decision becomes part of the pattern the system references the next time a similar product comes through, creating a feedback loop that keeps recommendations aligned with your organization’s classification standards and risk tolerance.

There is a reality that every global trade professional knows all too well: managing thousands of SKUs across 220+ countries while manually researching fragmented government sources. As Blair Kellum, Senior Implementation Consultant at Thomson Reuters Professional Services, explains, “By integrating CoCounsel into our global classification module, we are significantly reducing the manual effort required to navigate complex tariff schedules. This shift allows your team to achieve precise product classifications with greater speed and less administrative burden,” Kellum notes. The system compares key attributes to identify comparable products and classifications, helping professionals make informed decisions quickly.

The AI provides multiple classification suggestions to guide decision-making. For example, when classifying a brake pad, the system might suggest two options: one for an automobile, presented as the stronger match based on tariff schedule structure, GRI logic, and your organization’s own classification history, and another for a high-speed train, presented as the weaker alternative given that same evidence. This structured reasoning helps professionals see why one option is favored and evaluate each recommendation accordingly.

Research acceleration with authoritative sources

The Interactive Classification Research feature, powered by CoCounsel, provides direct access to enhanced tariff schedules, WCO notes, and binding rulings. This systematic application of General Rules of Interpretation uses trusted Thomson Reuters content to deliver comprehensive research capabilities within your existing workflow.

The AI HS code classification software automatically handles expired HS codes, offering valid replacement suggestions and ensuring continuity in your classification database. This eliminates the risk of using outdated codes while maintaining classification accuracy across regulatory changes.

For regulatory questions that extend beyond a single classification decision (a new tariff action, an executive order, an FTA eligibility question), ONESOURCE Global Trade Research powered by CoCounsel extends that same trusted content foundation across the full body of trade regulation, returning a cited answer in seconds rather than a stack of documents to work through.

Automated audit readiness

Every classification decision generates transparent rationales linking to precise regulatory documentation. The system provides comprehensive audit trails with GRI rule citations, tariff notes, and binding rulings, while historical decisions are recorded for compliance review and audit purposes. That audit trail with regulatory citations is what supports a “reasonable care” defense when CBP asks how a classification decision was made.

This automated documentation significantly reduces the burden and time needed for customs audit preparation, giving trade teams confidence in their classification defensibility.

Real-world impact: From busywork to strategic value

Efficiency gains that matter

Early implementations show 50% reduction in time spent on HS classification tasks. The system helps teams move through high-volume classification needs more efficiently, freeing experts for strategic trade initiatives rather than repetitive classification work.

As the system learns from user decisions, classification accuracy improves over time. “One of the most significant advantages of this model is its capacity for continuous learning,” notes Kellum. “By leveraging the classification decisions made by the professionals in your classification department, the AI identifies patterns and refines its accuracy over time. You are essentially digitizing the collective expertise of your team into a growing intelligence layer tailored to the SKUs relevant to your business.”

This means when you input “brake pad,” and you are an auto-manufacturer, the system won’t suggest a train brake pad classification because it has learned your business context and product patterns, offering more tailored classifications specific to your industry.

Confidence in complex decisions

Multiple predictions with detailed rationales support informed decision-making. Classification suggestions, prioritized by the strength of supporting evidence, help professionals choose the most appropriate classifications, while built-in handling of expired HS codes provides automatic suggestions for valid replacements.

The AI HS code classification software demonstrates sophisticated understanding by parsing through product acronyms and brief descriptions to understand the true intent behind each classification request, ensuring that nuanced decisions receive appropriate context and support.

Consistency across global operations

Centralized classification capabilities maintain global consistency while standardized product descriptions improve data quality across ERP systems. The new classification queue system provides a workflow-driven approach where teams can efficiently move from unclassified products through validation and final approval stages.

As Maldonado explains: “The new global classification queues feel like a true work-cadence process. You start by identifying unclassified or HS-invalid products that need correction. As classification progresses, additional queues, such as Tariff Changes, HS Suggestions, and Classification Audit, come into play. This structured workflow makes it easy for trade compliance teams to collaborate and move efficiently from one classification queue to the next.”

The Thomson Reuters advantage: Trusted content meets classification technology

Unmatched global trade content foundation

Thomson Reuters maintains 150+ researchers monitoring 1,300+ sources across 220+ countries and territories, delivering 155M+ regulatory updates annually with same-day content updates. This comprehensive coverage includes Federal Register notices, executive orders, and CBP guidance, all integrated into the AI’s knowledge base.

Domain-specific AI that understands trade

Purpose-built for global trade compliance rather than generic search, the system demonstrates deep understanding of HTS codes, duty drawback, customs procedures, and cross-border nuances. Integration within the trusted ONESOURCE ecosystem provides comprehensive workflow management.

That same content foundation extends beyond classification. It’s also what powers ONESOURCE Global Trade Research powered by CoCounsel, standing up to a real-world accuracy test most AI tools never face.

Proven accuracy and reliability

Every answer includes direct citations to authoritative sources for verification. This transparency allows compliance teams to validate and audit AI recommendations with confidence, maintaining the human-in-the-loop approach essential for regulatory compliance.

The system provides detailed breakdowns showing the complete classification hierarchy, from the initial two-digit chapter through the full ten-digit classification, with explanatory text helping users understand the logic behind each level of classification.

Why trustworthy AI starts with trustworthy sources

Classification and research both depend on the same non-negotiable requirement: the AI’s answer has to be traceable to a real regulatory source, every time. That standard is what separates ONESOURCE Global Trade Research powered by CoCounsel, and Global Classification alongside it, from generic AI tools that generate plausible-sounding answers with no way to confirm them. It’s the difference between AI product classification built for trade compliance and a general-purpose chatbot repurposed for it.

What makes Thomson Reuters trade content authoritative

Authority starts with depth and currency, not just volume. That same content foundation, described above, is vetted and structured by editorial teams who specialize in trade regulation, not scraped from the open web, and it spans Federal Register notices, CBP CSMS messages, executive orders, binding rulings, and the Harmonized Tariff Schedule across all 99 chapters. Global Trade Research’s knowledge base refreshes as regulations change, with average update times under one business day.

Cited answers compliance professionals can verify

ONESOURCE Global Trade Research powered by CoCounsel doesn’t return a summary you have to take on faith. It returns a cited answer, synthesized from that vetted source base, that points back to the specific regulatory document behind it. Ask about current tariff rates on a given HTS subheading, FTA eligibility for a product line, or what a new executive order changes for a supply chain, and the system delivers a direct, sourced answer in seconds rather than the hours a manual multi-source search would take.

Why traceability to primary sources is non-negotiable

Compliance teams don’t work from summaries. They work from the underlying regulation, because that’s what holds up when CBP asks a question. An AI tool that can’t show its work doesn’t fit that workflow, no matter how fluent its answers sound. Citing sources isn’t a feature bolted onto Global Trade Research after the fact; it’s the only way an AI tool can align with how compliance professionals are already trained to verify their own work: tracing every conclusion back to the primary document that supports it.

Proof in the highest-stakes exam in the field

The clearest test of that trust standard is the U.S. Customs Broker License Exam, one of the toughest professional licensing exams in trade, covering thousands of pages of CBP regulations, directives, rulings, and procedures, with pass rates historically as low as 2%. Global Trade Research has passed all six publicly available exams administered over the last three years (18 total runs, with a mean score above 84%), including through two years of significant tariff and regulatory change.

“Trade teams are being asked to become strategic partners rather than operational functionaries. They need tools that can provide instant, accurate regulatory intelligence to support real-time business decisions.” — Ray Grove, Head of Product, Global Tax and Trade, Thomson Reuters

This is Fiduciary-Grade AI™ in practice: every answer traceable, every source authoritative, built to support decisions where being wrong carries real professional and financial consequence.

Your next step: Build a faster, more defensible trade compliance workflow

See how ONESOURCE Global Classification and Global Trade Research, powered by CoCounsel, work together inside your existing workflow, pairing defensible classification decisions with instant, cited regulatory research as a single AI trade compliance software workflow. Request a personalized demonstration showing real classification and research scenarios relevant to your industry. Discover how your team can reclaim hours for strategic trade optimization while maintaining the accuracy and defensibility required for global compliance.

Connect with Thomson Reuters to explore implementation tailored to your specific industry needs and begin your journey toward efficient product classification and faster, more defensible regulatory research.