Using AI to Prepare the Answer to a Complaint
Artificial intelligence tools can help you get work done faster and better at every stage of a lawsuit – a topic I have been addressing in an ongoing series of posts. Today’s post examines how to use AI when preparing answers to complaints.
Earlier posts in the series are:
- AI Across the Life of a Lawsuit
- Using AI to Prepare Complaints: Part 1, The Complaint
- Using AI to Prepare Complaints: Part 2, The AI
Elements of an Answer
A lawsuit typically starts when the party bringing the case – the plaintiff – serves and files a summons and complaint.
The party against whom the case has been brought – the defendant – then needs to respond. Generally, the defendant has three options:
- Answer the complaint,
- Move to have a portion of the case dismissed and answer the remaining allegations, or
- Move to have the case dismissed in its entirety.
In its answer, the defendant should:
- State in short and plain terms its defenses to each claim asserted against it,
- Admit or deny each allegation asserted against it by the plaintiff, and
- When denying an allegation, fairly respond to the substance of the allegation.
When responding to a specific allegation, the defendant has several choices. These include:
- Admit the entire allegation,
- Deny the entire allegation,
- Admin the part of the allegation that is true and deny the rest, or
- State that the defendant lacks knowledge or information sufficient to form a belief about the truth of an allegation – an action that has the effect of a denial.
In its answer, the defendant also may assert affirmative defenses. An affirmative defense is “a defense in which the defendant introduces evidence, which, if found to be credible, will negate criminal liability or civil liability, even if it is proven that the defendant committed the alleged acts.” (Affirmative Defense, Cornell Law School Legal Information Institute Wex.) Affirmative defenses can include, for example, accord and satisfaction, arbitration and award, assumption of risk, contributory negligence, duress, estoppel, failure of consideration, fraud, illegality, injury by fellow servant, laches, license, payment, release, res judicata, statute of frauds, statute of limitations, and waiver.
Duty to Perform Reasonable Inquiry
Answers need to be signed by an attorney of record (or by the defendant if that party is not represented by counsel). By signing the answer, the attorney represents to the court that the attorney has performed a reasonable inquiry and from that inquiry has determined that:
- The factual contentions asserted in the answer have or likely will have evidentiary support, and
- The denials of factual contentions asserted in the answer either are warranted on the evidence or, if specifically so identified, are reasonably based on belief or a lack of information.
An attorney who signs a complaint but who did not perform a reasonable inquiry and make the required determinations faces potential sanctions.
Performing a Reasonable Inquiry
For each allegation, you should attempt to assess whether the allegation appears to be accurate.
Sometimes you don’t have access to the information needed to make this determination. When that happens, you typically respond with some variation of “lack knowledge or information sufficient to form a belief about the truth of an allegation”.
Sometimes your client can provide the information you need to respond.
Sometimes, however, you need to do some digging. If you have reasonable access to potentially pertinent electronically stored information (ESI), you can draw on that information. You might know exactly where to go for the needed information. You might be able to locate the necessary content using text or Boolean searches.
You might need – or want – to turn to more powerful capabilities. In the rest of this post, we will take a look at some of the many AI-powered capabilities you might be able to use to help determine how to respond to allegations asserted in a complaint.
Often, allegations are about communications. These communications might be, for example:
- Between the parties,
- Between one of the parties and a person or organization not party to the suit,
- Between individuals associated with a party, such as employees of a defendant company, or
- Between individuals or organizations not party to the suit
An example of an allegation referring to a communication between a party and a person not party to the suit is:
As a defendant responding to an answer, you probably do not yet have information from the plaintiff about calls made or received by the plaintiff. Later in the case you might get logs of phone calls, calendar entries, notes from the doctor’s office – but most likely not yet.
Depending on what ESI you already have, you might be able to locate communications that could help you determine how to respond to the allegations in this paragraph.
A powerful starting point can be a communications map. Start with what you know and explore from there. With this example, you might begin by searching for Dr. John Brown as an individual. As the previous paragraph in the complaint identifies the hospital where Dr. Brown worked, you could try searching for any messages to or from anyone at that hospital.
To show what that might look like, I searched the Enron data for “Brown”. Of course, Dr. John Brown did not show up there, but I did find a “Kimberly Brown”. You can begin to see who this Brown communicated with and how much. If you saw something useful, you could drill deeper.
For more about communications maps, go to Visualize: Analyzing Connections Between Communications
Concept Searching and Cluster Wheels
With the right AI tools, you can search for concepts and then use tools such as cluster wheels to find high-level concepts quickly, drill in for greater details, and ultimately get to potentially key messages and other content.
In our example, the lawsuit is about an allegedly defective product called an “inferior vena cara filter”. The 37-page complaint contains the word “filter” 166 times and the acronym “IVC” 109 times. To respond adequately to the complaint, you ought to search the available ESI for references to IVC filters.
A while back, we loaded the contents of the English version of Wikipedia into Brainspace. I went to that database, just because I thought there might be something related to IVC filters, and started with a simple concept search using the phrase “inferior vena cara filter”:
From “additional concepts”, I added “inferior vena cara” to the top concepts:
And navigated to the cluster wheel to begin to dig into the results:
For more about concept searching and the cluster wheel, go to 11 Reasons Lawyers Love Reveal's Brainspace Cluster Wheel.
Explore Relationships Between Concepts
Another way to explore the relationships between concepts in your data is via an interactive visual display such as the Brain Explorer.
From my concept search, I can click on “View Brain Explorer”.
Once I am in the Brain Explorer, I can quickly change the weight on a key concept:
Now the “inferior vena cara” concept is required, which means that the term must be in the concept search results:
I can ask the system to show me sub-concepts, by clicking on the main concept:
I can drill down into additional levels of sub-concepts, add concepts from the “Additional Concepts” section on the right, and ultimately look at individual files.
For more information about the Brain Explorer and how to use it, go to Introducing the Brain Explorer.
Additional AI-Driven Capabilities
Depending on the platform you use, there can be many additional AI-driven capabilities you can use to help formulate response to allegations, determine whether there appears to be support for affirmative defenses, or identify information that can help you decide what other actions to take.
These might include:
- Using AI Models to jump-start the process, pushing potentially useful files to the front of the line. See What Is An AI Model? and Layering Legal AI Models for Faster Insights.
- Using entity extraction to learn more about individuals and organizations, including email addresses and pseudonyms used, positions help, concepts discussed, people communicated with, and people who discussed similar entities. See Getting to Know You: Entity Extraction in Action.
- Using sentiment to help figure out who did what, when, where, how, and most notably why. See Getting Sentimental: Using Emotional Signals in eDiscovery.
- Using stylometry to find communications that may contain indica of fraud. See Stylometry and the Fraud Triangle.
- Deploying Active Learning to “find more like this”. See How Important is Active Learning for eDiscovery?
- Amping up TAR with the use of High Precision Classification to more quickly and accurately find similar content. See Legal Document Review's New BFF: High Precision Active Learning.
- Drawing on image recognition and labeling to find important content in pictures. See AI Image Recognition: The eDiscovery Feature You Didn't Know Existed, and Testing the Efficacy of Image Labeling.
Try AI for Your Next Answer
With the right AI-driven platform, you can prepare answers better and faster, confident that you are serving your client well.
If your organization is interested in learning more about making more effective use of AI across the life of a lawsuit and find out how Reveal uses AI as an integral part of its AI-powered end-to-end legal document review platform, please contact us.