
Sign up to save your podcasts
Or


In this companion Field Note to Episode 008, Ron walks through the practical steps lawyers can take to implement the emerging AI discovery standards discussed in Jeffries v. Harsco. He breaks the process into a simple three-legged stool: choosing the right vendor, properly configuring the tool, and handling the lawyer-side workflow and documentation needed to make AI use in discovery more defensible.
In Episode 008, I talked about the emerging legal standard for using AI in discovery and why lawyers need more certainty in this area. In this Field Note, I focus on the practical question:
How do you actually implement it?
The answer, in my view, is a simple framework:
The Three-Legged StoolTo use AI in discovery in a way that is more defensible, you need all three of these in place:
1) The Vendor Leg
Start with the right environment. In practice, that usually means an enterprise-level AI tool — one that operates in a closed system, does not train on client data, provides a secure environment, allows deletion, and gives you a way to define who has access.
2) The IT / Configuration Leg
Buying the right tool is not enough. You also need to configure it correctly. In this Field Note, I explain a practical workflow for doing that:
3) The Lawyer Leg
This is where the legal workflow becomes defensible. I walk through the downloadable documents that can help lawyers operationalize the standard:
The downloadable materials discussed in this Field Note are available here:
https://lawyeraitoolkit.com/deliverables
That page currently includes:
You do not need perfection. But you do need all three legs of the stool.
If you have:
…then you are in a much stronger position to explain and defend your use of AI in discovery.
Share ThisIf you know a lawyer who is:
send them Episode 008, this Field Note, and the Deliverables page.
Because this is exactly the kind of issue where certainty matters.
By Ron DrescherIn this companion Field Note to Episode 008, Ron walks through the practical steps lawyers can take to implement the emerging AI discovery standards discussed in Jeffries v. Harsco. He breaks the process into a simple three-legged stool: choosing the right vendor, properly configuring the tool, and handling the lawyer-side workflow and documentation needed to make AI use in discovery more defensible.
In Episode 008, I talked about the emerging legal standard for using AI in discovery and why lawyers need more certainty in this area. In this Field Note, I focus on the practical question:
How do you actually implement it?
The answer, in my view, is a simple framework:
The Three-Legged StoolTo use AI in discovery in a way that is more defensible, you need all three of these in place:
1) The Vendor Leg
Start with the right environment. In practice, that usually means an enterprise-level AI tool — one that operates in a closed system, does not train on client data, provides a secure environment, allows deletion, and gives you a way to define who has access.
2) The IT / Configuration Leg
Buying the right tool is not enough. You also need to configure it correctly. In this Field Note, I explain a practical workflow for doing that:
3) The Lawyer Leg
This is where the legal workflow becomes defensible. I walk through the downloadable documents that can help lawyers operationalize the standard:
The downloadable materials discussed in this Field Note are available here:
https://lawyeraitoolkit.com/deliverables
That page currently includes:
You do not need perfection. But you do need all three legs of the stool.
If you have:
…then you are in a much stronger position to explain and defend your use of AI in discovery.
Share ThisIf you know a lawyer who is:
send them Episode 008, this Field Note, and the Deliverables page.
Because this is exactly the kind of issue where certainty matters.