Salient Logo
Salient Logo

Strategic use of AI and device imaging yields results within 30-day fraud investigation deadline

A large petrochemical/energy firm approach Salient to assist in the identification of procurement fraud/collusion/kickbacks between staff and vendors in their African operations.

Client Sector: Large Petrochemical/Energy Firm
Technology Used: AI eDiscovery

The Challenge

Working around a 30-day investigation time limit

A large petrochemical/energy firm approach Salient to assist in the identification of procurement fraud/collusion/kickbacks between staff and vendors in their African operations. It was likely that the investigation would require in-country action, which would trigger the locally legislated 30-day time limit for the investigation’s completion.

Our Solution

Phase 1: AI-driven forensic analytics for preliminary suspect identification

To limit the time required to complete the investigation after triggering the local 30-day investigation time limit, we performed preliminary forensic analytics to narrow down the suspect list to the most likely participants.

First, we used computer assisted audit techniques (CAATs) to extract procurement and vendor information. We then worked to understand the internal controls and risks at play. From here, we ran both custom and database forensic analytics routines against the data to find procedural outliers, exceptions and other red flags.

Investigators then discussed these red flags with the client, identifying six key suspects for further investigation.

Phase 2: Onsite device imaging and accelerated AI-driven data processing/review

Going onsite, in-country, the Salient team imaged all six suspects’ laptops and mobile devices. This information was then processed into our eDiscovery platform, where we used a combination of AI/continuous active learning and dashboard reporting to accelerate review.

Outcome

Actionable evidence

The investigation was successfully completed within the 30-day time limit, enabling the client to legally act on their findings.

Summary

Strategic application of AI enables us to perform targeted in-country imaging of laptops and mobile phones belonging to identified procurement fraud/collusion suspects. This information is then processed and reviewed on our eDiscovery platform (leveraging AI/continuous active learning and reporting) to surface results within the legislated timeframe, thereby enabling our client to act on their findings.

Challenges:

  • 30-day legislated time-limit for local investigation

Results:

  • Phase 1:
    • Suspicious/unusual activities flagged
    • Confirmed suspects identified for further investigation
  • Phase 2:
    • 6x Suspects’ laptops and mobile phones imaged (in-country)
    • AI used to accelerate processing and review of imaged data
    • Dashboards created to assist in prioritising investigation
    • Investigation completed within local legislation’s 30-day time limit