United KingdomPublic sector2017 – ongoingSerious Fraud Office
OpenText Axcelerate machine-learning document review
The Serious Fraud Office adopted machine-learning document review, then had to correct omitted search results
Problem
Fraud investigations required the Serious Fraud Office to examine large volumes of digital evidence. Investigators needed to identify relevant material without reading every document in sequence.
Approach
The SFO began live Axcelerate pilots in June 2017. The platform used machine learning to categorize documents and prioritize material for human review. A later independent review described a configuration problem that omitted some documents from searches and required reconfiguration and reanalysis.
Results
No published figures
Takeaway
Prioritizing documents still left disclosure decisions with human reviewers. The configuration problem also required checks on whether searches covered the underlying material.
Categories
- Industry
- Public sector
- Business function
- Finance, risk and legal
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Vendor solution
- Evidence
- Vendor case study
Sources
Source summaries
OpenText reported successful live pilots beginning in June 2017 and preparations to manage new cases through Axcelerate. It described machine-learning document analysis but did not provide a measured time-saving result.
The independent review described Axcelerate's use to prioritize material for human disclosure review. It also recorded a configuration issue that omitted documents from searches and the resulting remediation work.
Full details are in the original sources.