How AI Supports eDiscovery Document Review Part 1

Everybody gets it – AI is a game changer when it comes to eDiscovery document review. Legal teams regularly use AI for review organization, culling, and hot doc identification. Today, tech-savvy attorneys use AI in other helpful ways, with chat, image, and privilege review leading the list. 

Short Message Data 

Address its prevalence 

The medium of global communication has fundamentally changed. People now generate more data in chat and collaborative tools like Teams, Slack, Bloomberg, SMS, and WhatsApp than in email. As a result, the incidence frequency of short message format data in discovery sets has steadily increased over the past 2-3 years. And experts expect chat to surpass email as the dominant communication format in the next few years. But short message data creates unique challenges for eDiscovery.

Put them in context 

Since each chat message is a short snippet of the conversation, you need to see those individual messages in context to understand the stories they tell. Lineal’s ChatCraft keeps messages in their original native form – as conversations – so you can review threads just as you would if you were in Teams or a chat app. By applying CAL to these conversations, you can quickly recognize the main topics and identify relevant interactions. Message context increases accuracy and streamlines review. 

Image Data 

Use words to search for non-words 

Dealing with images in document sets is challenging. They take time to load, have no subject or meaningful metadata, and are impossible to categorize. Lineal offers the solution – AI-categorized images. It analyses the components of the images and assigns each to applicable categories. Take handwriting, for example. Lineal Images automatically identifies all images containing handwriting. It also categorizes documents with signatures, printed text, and colored ink. Using these categories, you can quickly identify all documents signed in red ink and view them in an intuitive gallery of thumbnails. Beyond finding what you are looking for, you can also identify what you aren’t looking for. 


Eliminate irrelevant images 

So many images in a data set are irrelevant. Most discovery collections are riddled with logos, gifs, and other typically non-responsive material. Lineal Images reduces the workload for reviewers by providing categories. You can quickly scan and remove non-responsive content en masse to reduce review time and related costs.  


AI is being applied to today’s data to improve how we review short messages and images. It is also changing the way we identify and classify Privileged content. Keep an eye out for our next blog, where we will look at that use case.