Log in

The Future of Discovery and Data Analytics in Divorce

6 Oct 2020 10:00 AM | Kirsten Peterson (Administrator)

Daniel R. Roche, CPA/ABV, ASA and Stefanie A. Jedra, CPA

During our presentation “The Future of Discovery and Data Analytics in Divorce” we shared how our team at Marcum has utilized artificial intelligence (“AI”) and machine learning to extract data and significantly reduce the time spent on data entry. We also discussed how we seek to continually improve the depth and quality of our analysis using data analytics.

There were some questions regarding Marcum’s processes related to data extraction, the review / analysis of the extracted data and how conclusions derived using extracted data would be perceived in the court room. These questions lead to a discussion about whether other companies and/or accounting firms were utilizing similar technology. This blog post seeks to address these questions and provide some additional background on the reach of AI in the accounting, finance, and legal communities.

The Data Extraction Process

In forensic accounting engagements, data is generally required to be entered into a useable electronic format – preferably Microsoft Excel.  From there, the data is analyzed and conclusions presented in a summary format to educate the trier of fact about the conclusions reached. 

Historically, this data was manually entered by administrative assistants / paraprofessionals and, depending on the amount of data, the process could take weeks or months to complete.  Not only was the work tedious and time consuming, but these administrative assistants / paraprofessionals often had other competing tasks to accomplish.  Utilizing computer processes and AI to extract data allows us to get the needed information into a useable format quicker and more accurately than the manual processes historically utilized.  This technology also allows the administrative assistants / paraprofessionals to be more focused on their other job responsibilities.

Regardless of the data entry method (human vs. computer) Marcum’s Quality Control procedures with respect to vetting the underlying data and review of the analyses completed remain unchanged. With the ability to extract larger volumes of data, we have been able to more thoroughly understand the universe of information affecting our conclusions, while reducing the cost to our clients. This has enabled us to provide conclusions that are more informed and meticulously supported.

AI in the Courtroom

There appear to be many court cases involving the extraction of data; however, these cases involve the extraction of data from cellphones using software that requires specialized or technical knowledge to get the data in a format useable / presentable in the court room. Accurately and reliably extracting the data from a cell phone requires technical knowledge based on training and experience that a lay person does not possess. It is not possible for a lay person to verify the resulting extracted data, which has caused questions about the reliability of the data under the rules of evidence.

Unlike the extraction of data from cellphones, it would be possible for a lay person to verify each line of data extracted from any bank statement, general ledger, or other document. As demonstrated during our presentation, a column populated with the bates numbers, pages numbers, or another identifier is included within the extracted data and enables a lay person to easily source the information back to the source document.

AI in the Accounting, Finance and Legal Communities

Artificial Intelligence is not new in the accounting, finance, and legal communities; however the use of AI has become further reaching over the last few years. In June 2016, JP Morgan launched “COiN” which stands for Contract Intelligence. Using image recognition software the program reviews commercial loan agreements, which the bank indicated once took lawyers 360,000 hours of work each year! In JP Morgan’s 2016 annual report, the bank highlights the success of this program by noting the capabilities have “far-reaching implications considering that approximately 80% of loan serving errors today are due to contract interpretation errors.”

Even more impressive, in JP Morgan’s 2019 annual report, the bank indicated that they have “used technology and machine learning to reduce fraud losses in the credit card business by 50%”. In the same report, the bank stated that they plan to continue to “evaluate emerging technologies and reshape [their] approach to data to bring the power of artificial intelligence and machine learning to all [their] businesses”.

It was great seeing everyone at our presentation.  Please reach out to us if you have any questions on this topic.

 

Sources: 

https://www.supremecourt.gov/DocketPDF/18/18-8600/92968/20190322193407473_McLeod%20CERT%20Final1.pdf

https://www.bloomberg.com/news/articles/2017-02-28/jpmorgan-marshals-an-army-of-developers-to-automate-high-finance

https://www.jpmorganchase.com/corporate/investor-relations/document/2016-annualreport.pdf

https://www.jpmorganchase.com/corporate/investor-relations/document/annualreport-2019.pdf

Ibid. 


AAML New Jersey

215 East Ridgewood Ave, Suite 201

Ridgewood, NJ 07450

CONTACT US

Office: (201) 445-7007

Email: contact@aamlnj.org
Powered by Wild Apricot Membership Software