Case Study
IQ.AI Uncovers Insider Risk and Intellectual Property Theft Case
When an anonymous tip alleged that a company’s former employee was breaking contract and stealing intellectual property, the organisation needed to quickly determine whether the allegations were true, and if so, the evidence available to form a legal case. FTI Technology worked with the organisation’s law firm to apply IQ.AI by FTI Technology™ for rapid fact-finding and analysis across numerous data sources.
Situation
Based on the tip, the former employee was suspected of recruiting existing staff members and clients in pursuit of starting a competitive business, which qualified as violations of the departure agreement. Additional information suggested that the employee had also stolen IP prior to leaving the company. With each day that passed without action, there was an increasing possibility of the individual misappropriating more employees, clients and information.
The organisation conducted an internal investigation, which did not uncover any evidence to support a case. Given the time pressures and sensitivity of the matter, the client’s law firm needed to quickly determine whether they could find the necessary information to form a case against the individual. This would require extensive analysis of more than 100,000 documents and activities associated with the former employee and the individual’s core team, spanning email, Microsoft 365 sources, chat messages and Microsoft Copilot prompts and logs.
Our Role
FTI Technology was engaged to conduct the investigation and fact-finding process within a three-day timeframe, so the general counsel could quickly decide the next course of action.
Even with machine learning technologies, reviewing this volume of documents would typically require multiple weeks of work. FTI Technology applied IQ.AI for Investigations, initially across a small set of the documents, to help guide the direction of the case, provide workflow flexibility and surface key information early on.
With guidance and oversight from the legal team and FTI Technology’s investigative experts, the work included:
- Prompt refinement to hone accuracy across findings.
- Application of generative AI solutions alongside continuous active learning models to conduct entity analysis and feed that into prompt guidance, allowing the models to simultaneously identify suspicious activity between work and personal email and other communications tools.
- Discovery of coded conversations occurring between key individuals and clients during the former employee’s garden leave — in addition to chats indicating intention to switch to off-channel applications — all which suggested irregular activity and violation of the former employee’s contract.
- Confirmation of attempts by the former employee to recruit existing employees.
- Identification and examination of additional documents that the organisation didn’t know it needed to consider, including receipts and submission of expenses that occurred during periods employees were on paid leave, flagging additional suspicious behaviour.
- Detection of large volumes of company data, including confidential pricing information, being transferred to personal cloud drives.
- Support for the ongoing investigation, including collection and analysis of key custodian mobile devices and cloud drives, which revealed further evidence of violations and IP theft.
Our Impact
FTI Technology delivered a fast and flexible AI-powered solution to find key information under time pressure. The ability to uncover evidence within a matter of days provided the client with the evidence needed to initiate legal action against the former employee and mitigate further IP loss.
Identification of 120 key pieces of electronic evidence, extracted from a complex set of more than 100,000 documents, supported critical next steps in legal proceedings, all within the client’s time and budget constraints.
FTI Technology’s experts are supporting the ongoing investigation, providing digital forensics, expertise and IQ.AI solutions as the case progresses.