From AI to Action: Advanced Workflows for Smarter Risk Management

The global maritime risk landscape has undergone a profound transformation since the COVID-19 pandemic began in late 2019, necessitating advanced risk management capabilities and the means to configure risk parameters for trade resiliency and security. 

Geopolitical disruptions are more frequent than ever before, and deceptive shipping practices (DSPs) by bad actors have become more sophisticated and opaque. The challenge is to maintain an adaptive and dynamic approach to risk management without creating time-consuming and costly false positives – an approach that can quickly identify new risks, define them as such, generate leads, and monitor and respond to them.

Whether you are focusing on growing your business in light of recurring trade disruptions and regulatory penalties, or working to defend your borders from security threats, risk management is a must-have. 

Risk mitigation can help you stay ahead of emerging global threats to quickly identify maritime anomalies, better strategize, and allocate and deploy resources more effectively and efficiently. 

For stakeholders across the shipping supply chain, AI-powered risk management is the only way to navigate shifting trade flows and increasingly complex sanctions evasions, grow their business, and stay ahead of the competition.

A one-size-fits-all approach to risk management is counterproductive and counterintuitive at a time when it is possible to pinpoint the exact risks organizations care about most at any given time, anywhere in the world.

Shipping and security stakeholders can ensure that their risk management strategies align with their operational and strategic risk mitigation objectives by adopting a configurable approach. 

Windward sets up its users for success by equipping them with the most advanced and best-in-class lead generation and risk management tools. This includes two central tenets: 

• Customization of queries for enhanced precision and alignment with the organization’s interests and risk parameters
• Configuration of relevant and organization-specific risk profiles for ongoing lead generation and consistent monitoring

Put differently: you can search for exactly what you’re looking for to minimize false positives, generate quality leads for further investigation, and grow your business, while ensuring compliance. Then, it’s easy to determine the risk level that makes the most sense to you and your organization.

Windward recently (2024) released several capabilities that take search to the next level:

• Sequence Search: search for sequences, rather than a stand-alone activity, creating customized queries with sequential conditions between activities
• Ship-to-ship (STS) classification: a search criteria to distinguish between commodity, bunkering, and logistics meetings, for more accurate business insights and strategic decisions
• Second vessel criteria: second vessel querying capabilities to efficiently extract relevant data sets about who vessels of interest are engaging with

By searching for sequences of activities and zooming in on relevant STS meetings, users can boost the quality of leads they generate, and leverage sequential context to minimize false positives and ensure they focus on what matters most. 

To close the loop, Windward enables users to define and apply individual risk parameters that resonate with their specific needs, scenarios, industries, and teams – ensuring that future vessels with similar behavioral trends or characteristics are marked as either medium or high risk, or simply activating an indication for the behavioral pattern one is monitoring.

The above search tools are powerful on their own, but Windward took risk management to the next level with configurability and cutting-edge Gen AI tools. 

The industry’s first fully configurable risk solution, Organization Defined Risk (ODR), enables users to tailor behavioral indicators to meet their unique business and risk needs. Indicators can be applied to both behavioral and static screening data sets.

Customization is important for increasing accuracy and efficiency, but today’s stakeholders can only select predefined settings or parameters. A truly configurable approach enables organizations to define and apply individual risk parameters that resonate with their specific needs.

To ensure investigation and screening standardization, accuracy, and efficiency – while saving time and resources – Windward created the industry’s first maritime Gen AI agent for vessel risk assessment and insight summaries.

One best practice enabled by advanced search features is to start with a vessel confirmed to have been involved in illicit activities or sanctions breaching. Investigating the vessel will reveal its activities before it was intercepted: where it came from, whether it made any stops or deviations, and if it engaged in deceptive shipping practices (DSPs) during its voyage. 

Illicit actors tend to mimic each other’s methods of evasion, patterns develop. Unveiling one bad actor’s method through reverse-engineering of its voyage is instrumental in shedding light on a possible pattern of operation and generating additional leads potentially engaged in similar activities.

The Sweet Miri vessel is a good example. After catching this bunkering tanker vessel with two million liters of stolen oil(!), Nigerian authorities were made aware of a pattern employed by the vessel: 

• A port call in Nigeria
• Going dark as it exited the Nigerian exclusive economic zone (EEZ)
•  Resuming transmission upon return 

Using Windward’s Sequence Search capability, authorities identified 43 vessels displaying a similar sequence of activities.

These 43 vessels engaging in a behavior known to be risky can be configured as either as a medium/high risk, with an Organization Defined Risk, adding another risk indicator that will automatically work for them 

Additionally, shipping stakeholders can ensure their fleet is not engaged with any of them and avoid possible exposure to sanctions with an additional ODR indication for when a member of its fleet is meeting one of these suspicious vessels. Law enforcement agencies or coast guards can safeguard their borders and EEZs from being exploited via oil theft and smuggling.

Sequence Search is a pivotal capability that makes reverse engineering possible, allowing organizations to leverage successful screenings or interceptions into a more nuanced and precise understanding of much larger patterns and trends. 

This approach is instrumental for securely growing and scaling your business, safely entering new markets, and taking advantage of new opportunities with minimal exposure.

As an analyst logs into the Windward Maritime AI™ platform at the start of their day, an alert pops up on their screen: an anomaly showing an increase in ship-to-ship operations of vessels marked high risk for illegal, unreported, and unregulated (IUU) fishing in the North West Pacific Ocean. 

Once the anomaly is detected, the analyst’s central concern is whether the increase in IUU fishing operations is violating their country’s jurisdiction, or violating its laws or regulations – and whether action is required. Using Windward’s Sequence Search, the analyst sets up the following query, looking for:

• A reefer engaging in a ship-to-ship meeting in this new hot zone The ship-to-ship meeting is classified as a commodity transfer
• The second vessel involved in the meeting is a fishing vessel, marked as high risk for IUU fishing
• The reefer then conducts a port call to offload its cargo

The query, utilizing Sequence Search, STS classification, and second vessel criteria, allows us to understand where these reefers called port and potentially offloaded their illegal catch. The results provide visibility into the operations of vessels possibly engaged in IUU fishing in a new hot zone, as well as into their support fleet. 

The query results reveal that 15 reefer vessels have engaged in ship-to-ship meetings with IUU fishing high risk vessels in the North West Pacific Ocean over the past year. Seven of these reefers are marked as high or moderate risk for IUU fishing, smuggling, or forced labor.

Sanction violations are growing more complex with every passing moment and in most cases, Winward uses a combination of behaviors to reveal potential attempts at sanctions circumvention – representing a possible risk of sanctions exposure for customers. 

Windward’s sanctions compliance model leverages multiple complex AI/ML models as building blocks. 

The following workflow demonstrates a combination of on-shore behavior, such as a port call, either while transmitting AIS or not, and a consecutive ship-to-ship operation. This sequence could indicate cargo exchanging hands along the way to its final destination.

Let’s use Russian crude oil as an example. An investigation can begin by looking for tankers longer than 180 meters that either conducted a port call or a dark activity in the Russian EEZ. This query currently identifies 613 vessels that potentially interacted with Russian vessels – too high a number to effectively and efficiently screen individually.

We save this target population as a “vessels of interest” list (VOI) and filter them even further into a workable list of interest. This is done by querying for vessels within our VOI list that 

• Engaged in  a ship-to-ship meeting that: 
  • Classified as a commodity transfer, or as “unclassified cargo”
  • Lasted over eight hours
  • The second vessel was a tanker

The results reveal that out of our initial 613 vessels, 80 vessels engaged in 106 ship-to-ship operations, all already flagged as high risk by Windward’s AI-based sanctions compliance and Russia risk models. These vessels can be saved into a second VOI list.

Rudimentary search capabilities are no match for the sophisticated methods applied by bad actors attempting to evade sanctions, smuggle goods, or any other illicit activity. Effective and impactful risk management relies on having the right tools to pinpoint and search for specific behaviors, patterns, events, and entities. Windward can help!