Getting value from investigations: how AI and forensic technologies can create tangible efficiencies

On 4 June 2026, Osborne Clarke and Forensic Risk Alliance (FRA) co-hosted a panel discussion as part of London International Disputes Week, exploring how organisations can extract real value from investigations and where technology, particularly AI, is changing the landscape. The panel brought together Osborne Clarke investigations and disputes partners Jane Park-Weir (head of corporate investigations) and Chris Wrigley (co-head of global compliance), Jon Fowler (head of FRA's global forensic technology team) and Alison Geary (head of litigation at Vodafone plc, but speaking in a personal capacity).

The discussion followed the full lifecycle of an investigation, from triage and scoping through to data analysis, interviews and reporting. The panel considered where AI and other technologies genuinely help, where they introduce risk, and how to maintain human judgement at the core of investigations.

The changing investigations landscape

Corporate investigations are becoming a more prominent and important aspect of business operations than ever before. Although the number of internal investigations being run as "business as usual" within organisations has increased markedly, the panel noted that experience indicates the quality of those investigations can be mixed, with some being highly sophisticated while others fail to get basic investigative steps right.

Matching this volume increase, technology has transformed how work gets done: predictive coding (once treated with caution) is now standard for document review and analysis, and the application of generative AI is seen as the natural next step. Client requests for proposals increasingly ask in detail about how firms intend to use technology, and specifically AI, to maximise efficiencies in both time and costs.

There can, however, be a tension between the expectations and reality of how technology and AI can help.

The panel observed that the prevalence of investigations means some businesses are starting to approach investigations as the answer to a question, rather than as a process to enable informed decision-making. In order to mitigate this risk, the initial scoping and design stage of investigations remains a key area to get right. The importance of human judgement at this stage cannot be overstated and the panel agreed that it was essential both to choose the right technology for the task in hand, and to know when technology cannot replace human expertise and analysis.

When to bring in external support

Many organisations have sophisticated investigations knowledge and experience in-house, along with a high level of familiarity with AI and technological tools. External support tends to be sought when an external threat emerges, such as regulatory scrutiny or litigation risk, or when distance and independence are needed to give the investigation credibility and defensibility.

The panel agreed that external counsel can continue to add value by bringing instant access to broader legal expertise across regulatory regimes and jurisdictions, helping to ensure the design and scope of an investigation is clear, objective and justifiable to stakeholders and regulators. They can also play a role in helping in-house teams deliver difficult messages to senior stakeholders.

AI in practice: opportunities and guardrails

AI is well-suited to handling large structured datasets and document review, capable of spotting patterns and anomalies quickly to support faster prioritisation of material. Successful use cases that the panel has seen include using AI to draft investigation process documents, support preparation for witness interviews and enable investigators to query and interrogate large document sets.

The panel agreed there are challenges in this context. Hallucinations remain a major concern, outputs must be checked and validated, and significant lawyer time can be spent developing and refining prompts. This means that the use of AI may create the need for a new type of work by the legal team rather than removing the need for lawyers to be involved; not every stakeholder in a business understands this distinction.

Investigations often involve "finding the narrative in chaos" and reconstructing complex stories from messy data. In this situation, AI can currently struggle with the nuance, context and meanings in data that can be crucial, such as, for instance, the use of a particular emoji to represent a state of mind.

The practical message is clear: organisations and their lawyers need to think carefully about why technological tools are being used and which tool is right for which job. When businesses or their stakeholders insist AI must be used, the right response is a conversation about "why" and "for what" at an early stage so that the most appropriate solutions are implemented from the start.

Where value can be found

Core sources of value in investigations include judgement, human engagement (including the ability to speak to people, test their recollections and read the room) and independence. These are factors which are difficult to put a price on but that are crucial to the output and credibility of an investigation.

The panel discussed the fact that using the right technological tools can save hours of human time and considered the likelihood of external legal and other advisers moving further towards value-based billing for investigations work instead of the billable hour. They agreed there was strong potential to use hybrid models mixing traditional hourly rates with value-based or fixed-fee elements.

Looking ahead

The next five years are expected to bring a period of steep learning about how and where AI and technological tools work best, even as those tools develop. This will be coupled with an increase in investigations and pressure to conduct them, and a growing regulatory focus on how companies are using AI in their decision making.

The panel anticipated that the use of AI itself would come under investigation, including as to the appropriateness of AI-driven actions. There remains great opportunity and value in combining existing investigative skills with new technologies, but the emphasis must remain on deploying such technologies with rigour and judgement.

Practical takeaways for in-house legal and compliance

• Invest time in scoping. Use the triage stage to set clear objectives, understand context and define the right questions before deploying technology.
• Use AI selectively and transparently. Do not assume AI must be used in every investigation; be ready to explain when it is appropriate, how it is used and what safeguards are in place.
• Engage external expertise early. Consider external counsel, forensic and other specialists where independence, cross-border issues and/or specialist technical skills are required.
• Plan for mobile and messaging data. Anticipate the practical and legal challenges in accessing phones and encrypted apps, and factor this into timelines and scoping.
• Maintain human judgement. Even where AI supports document review, interview planning or reporting, ensure lawyers remain close to the evidence and actively challenge AI outputs.
• Document a defensible process. For each investigation, maintain a record of how tools were chosen, how AI was used and how outputs were validated, in anticipation of future scrutiny.
• Future pricing models. Consider hybrid pricing that reflects both AI-enabled efficiencies and the value of independence, experience and human judgement.