Baltic Exchange is working with One Beyond to use AI to transform complex counterparty risk reports into clear, focused summaries.

By applying Anthropic’s AI technology to its Counterparty Risk Scan service, Baltic Exchange aims to help clients and members review extensive sanctions, political exposure, adverse media and other risk information more quickly and efficiently. The solution highlights the most recent and potentially significant findings for each returned entity, while directing users back to the supporting source records. It is designed to support efficient reviews without making compliance decisions or generating its own risk scores.

At a glance

Faster reviews

AI-generated summaries reduce the need to manually work through hundreds of pages of counterparty risk information.

Clearer risk information

The most recent and potentially significant findings are presented in a concise, structured and accessible format.

Anthropic-powered summaries

Anthropic’s AI technology interprets complex risk information and produces a focused summary for each returned entity.

People remain in control

The solution does not make compliance decisions or create its own risk scores. Users remain responsible for reviewing the evidence and making the final assessment.

Secure, separate processing

Each entity is processed independently, enabling results to be handled in parallel while reducing the risk of information being mixed or confused between entities.

About Baltic Exchange

Baltic Exchange is a leading global source of independent maritime market information, supporting the international shipping community through trusted data, benchmarks and professional services. As part of its wider offering, Baltic Exchange provides a chargeable Counterparty Risk Scan service that clients and members can use to support Know Your Customer (KYC) activities. KYC is the process of verifying the identity of an individual or organisation and assessing potential financial, legal or reputational risks before entering into, or continuing, a business relationship.

The Challenge

A single Counterparty Risk Scan can return multiple possible entities and hundreds of pages of information covering sanctions, politically exposed persons, adverse media, tax offences and other risk events.
Baltic Exchange’s clients and members must review this material to identify the most recent and potentially significant findings, before locating the supporting records required for further investigation.
Although this information is essential to effective due diligence, the volume and complexity of the reports can make the process extremely time-consuming. Important findings may be buried within extensive results, making it difficult for users to see quickly where their attention should be focused.
Baltic Exchange identified an opportunity to use AI to make this information more accessible and actionable – without allowing the technology to make compliance decisions on the user’s behalf.

Our Solution

Baltic Exchange partnered with One Beyond to explore how AI could enhance its Counterparty Risk Scan service and make large, complex reports faster and easier to review. One Beyond developed a workflow using Anthropic’s AI technology to produce a concise, structured summary for each entity returned by a scan. The summary highlights the most recent and potentially highest-risk findings and directs users back to the relevant supporting records. Each entity is processed separately. This allows multiple entities to be handled in parallel for greater efficiency, while reducing the risk of cross-contamination or confusion between their results. The solution does not make a compliance decision or generate an independent risk score. Its role is to organise and summarise the available information so that users can identify where to focus their attention while remaining fully responsible for the final assessment.

An iterative approach to better AI outputs

Developing an effective AI solution requires more than simply connecting a model to a source of information.
One Beyond has taken an iterative approach to prompt development, testing and output refinement. The summaries have been progressively improved to align with what Baltic Exchange considers most useful and efficient for its clients and members.
This collaborative process helps ensure the AI output reflects the practical needs of the people using the Counterparty Risk Scan service, presenting the right level of detail without losing access to the underlying evidence.

A closer look at features and benefits

AI-generated entity summaries – Turn extensive and complex counterparty risk information into concise, structured overviews.

Priority findings highlighted – Draw users’ attention to the most recent and potentially significant sanctions, political exposure, adverse media and other risk events.

Links to supporting records – Direct users back to the underlying source material so that findings can be checked and investigated.

Separate entity processing – Process every returned entity independently, reducing the risk of information being mixed between different results.

Parallel processing – Allow multiple entities to be processed simultaneously, supporting faster overall turnaround times.

Iterative prompt refinement – Continuously improve the relevance, structure and usefulness of the summaries based on Baltic Exchange’s requirements.

Human-controlled decisions – Support users with clearer information without making compliance decisions or generating new risk scores.

Reduced review time – Help clients and members spend less time navigating lengthy reports and more time investigating the findings that matter.

Enhanced client service – Add further value to Baltic Exchange’s chargeable Counterparty Risk Scan service through a faster and more accessible review experience.