How The Waldrep Company Uses AI

What AI never touches at The Waldrep Company

• No evidence, ever. No device extraction, image, log, exhibit, or finding from any case, active or closed, enters any AI system. AI is not a forensic tool here. It does not examine evidence, it does not form opinions, and it does not write findings.
• No case information. No personally identifiable information about any victim, suspect, defendant, witness, minor, or client goes into an AI system. No case names, numbers, dates, or locations. Case patterns used in training materials are fully generalized and synthetic.
• No privileged or client-confidential material. Attorney-client and work-product material stays out of AI systems entirely. Client-confidential business information enters an AI system only with the client's explicit authorization.
• No regulated data. Criminal justice information, law-enforcement-sensitive data, and export-controlled technical data stay in their approved environments. AI tools are not one of them.
• No expert opinions. Every opinion I sign is mine, formed from my own examination of the record. Nothing AI produces enters an evidentiary record or a court filing as analysis.

What we do use AI for

Business operations: research on public sources with every material claim verified against the original, drafting and structure work on marketing and training materials, scheduling, correspondence handling, and administration. Every AI-assisted output is reviewed by a person before it is used, and business-system integrations are confined to business operations, never casework.

The discipline underneath it

• AI work happens in business-tier accounts we administer, not personal accounts, and those accounts do not train on our content.
• If sensitive material ever reaches an AI workflow by mistake, work stops, the event is flagged, and the material is removed before anything proceeds.
• When a model or tool changes underneath us, we treat that as a reason to re-check the workflows that used it.

Why our line is stricter than yours needs to be

A working lab or litigation team can, with the right controls, use AI on more than we do: our red line sits where it does because expert testimony is the product here. When we design a governed AI program for a client, the client's counsel decides where the client's lines sit. Ours are published so you can hold us to them.

Questions