April 25, 2022

From AI to Neural Networks, Forensic Technology Looks to the Future

The demand for legal forensic technology continues to grow at a frenzied pace and the necessity of investing in those applications will increase along with it. At least one think tank predicts the global market for these services will expand from $10.9 billion in 2020 to $15.12 billion by 2025.

Meanwhile, millions of workers have had to deal with the real problem of pandemic lockdowns, making it more difficult for forensic technology professionals to physically secure electronic devices and collect information. With the growing adoption of remote work and the increasing demand for technology, the eDiscovery industry accelerated movement toward a higher acceptance of cloud-based evidence as technology improves enough to satisfy courts and regulators.

What that future will look like and when it will arrive is anyone’s guess, but there are key changes that are pushing practitioners to hybrid systems in the industry’s march toward platforms that all parties can agree upon and do not require physically handling digital media stored on devices.

Unprepared for the New Remote Work Reality

COVID-19 took everyone by surprise, of course, but for businesses that are focused on eDiscovery, it quickly became apparent that employees were now working on less secure networks, especially in some countries where broadband was simply subpar. Companies had to quickly work through those issues.

Hard drives and storage media —the familiar, tangible evidence sources of data relied on by courts and investigators — suddenly became difficult or impossible to physically collect and review. Vendors responded with tools to capture data and automate processes remotely. While the tools capture data, some of the data isn’t considered forensic evidence in court if not captured in a forensic manner (forensic images).

One of the solutions vendors provided in recent years was a way to retrieve data via the cloud while retaining the timestamp and metadata inherent in the original file, a step that may become more acceptable going forward.

Certainly, the tools are highly effective in capturing data remotely, allowing litigators and investigators to find red flags before they become a problem. Still, criminal or fraud evidence clearly require a high level of data security and very fast network connections, and the acceptance of cloud data retrieval is yet to be adopted in some jurisdictions.

Expect Forensic Technology to Evolve

As it stands, there is a variety of software platforms that help practitioners better sift mounds of documentation more easily.

Some tools allow forensic investigators to collect terabytes of email data from around the world, but reviewing it requires the use of some kind of Artificial Intelligence (AI) or active learning techniques to find patterns and trends in the data; otherwise, it may become too expensive and nearly impossible to complete reviews of collected information within a reasonable timeframe by employing only traditional methods (e.g., keyword search review). And while technology assisted reviews are now widely applied and accepted by legal professionals, some attorneys still prefer to rely on standard, familiar techniques of collection and linear review. However, recent developments and the more extended use of these forensic technologies proved they work and hugely benefit attorneys and forensic investigators with their e-Discovery obligations.

Here are some brief summaries of evolving technology in eDiscovery:

  • Artificial Intelligence: Already used to actively learn and detect data trends or anomalies in vast amounts of digital information, AI could further produce eDiscovery efficiencies across many forms of technology and documentation.
  • Technology Assisted Review (TAR): This is a kind of human-in-the-loop approach where, starting from an initial query by the user, ranking algorithms are continuously trained according to the relevance feedback from the user until a substantial number of the relevant documents are identified. This approach, named Continuous Active Learning (CAL), is more effective and more efficient than traditional e-discovery and systematic review practices, which typically consist of a mix of keyword search and manual review of the search results. TAR also include predictive coding, email threading and clustering concepts to help narrow down the reviewable set documents and pin-point red flags and relevant pieces of information.
  • Blockchain: Best known for its roll in cryptocurrencies, this technology stores digital information in a distributed database allowing for a secure record of transactions. Because the records cannot be changed, it’s a popular way to capture original metadata, but is unlikely to be the future solution in eDiscovery.
  • Cloud Services: Discovery processes can be thwarted because of the simple nature of cloud services. Where is the cloud data stored? Who has custody of the information? What controls are in place to secure the data? What regional or country restrictions exist to prevent transfer or even capture of information. What are the network limitations? Cloud services requires other technologies to help answer those questions affirmatively in the future.

Data Volume and Retention Are Only Going to Multiply

In the past, when email was the primary (and often only) unstructured digital documentation to process, a one-platform solution sufficed. As devices and types of data have grown, eDiscovery practitioners must employ a variety of technologies.

Consider the Internet of Things (IoT), the moniker assigned to describe all the data collected or distributed by devices as sophisticated as self-driving sensors in automobiles and as domestic as internet-connected refrigerators. How do you best review and account for the data explosion spun from millions of consumer and commercial device sensors and computing applications?

Will we ever see a one-stop eDiscovery technology platform? Perhaps, but in the meantime, new technologies are evolving to cover some gaps until that day.

Our A&M offices in Brazil, United Arab Emirates, Hong Kong, India, United Kingdom and the United States offer Relativity One, a global full-featured e-discovery application that’s widely adopted and considered the market leader. It provides flexible, custom reviews and quickly identifies issues during investigations and litigation. And it’s a more cost-effective way to host cloud-based data than proprietary systems because it works with Microsoft’s off-the-shelf Azure One.

Most likely, the future of eDiscovery will center around neural networks or quantum computing. Neural networks, as the name suggests, act similarly to our brains by developing artificial neurons, or nodes, to process information. It uses deep learning to correct mistakes as it goes, improving accuracy over time. Google Translate, Echo’s Alexa and Tesla’s self-driving systems all use this technology.

Quantum computing takes its name from the science discipline of physics and essentially removes the requirement of using binary bits (ones and zeros) to process or calculate data. Instead, it considers each input to be both a one or zero at the same time, similar to the way scientists have discovered particles can be in two places at the same time. Effectively, what that means is rather than submitting each input one after another, quantum computers and work with all at once and sweep away the wrong answers at the end, exponentially speeding up eDiscovery.

These technologies, and others we’ve not yet developed or applied to eDiscovery, may also be around the corner. Until then, litigators and investigators have decisions to make about the right tools to use on each case.

Investigators must not be afraid to use some of the new products to capture data remotely and should be on the lookout for changing court acceptance measures for some evidence.

Conclusion

For the immediate future, litigators and investigators should take a bespoke approach to eDiscovery, selecting from a portfolio of software tools to match the task and to scale up and down, depending on the data volume. This hybrid perspective allows investigators to apply technology to locate red flags and then dig deeper manually to uncover evidence.

eDiscovery will continue to evolve, investment in forensic technology will keep growing and the job for litigators and investigators will be more complex. Working with a spectrum of technology tools will help manage both the volume and processing time of data in your next case.

 

For more information, click here to contact us.

To read the article in The Oath, click here.

 

FOLLOW & CONNECT WITH A&M