What Prevaldi is, for this purpose
Prevaldi is a small company. When you use one of the free legal tools at prevaldi.com/legal, your input goes from your browser, through our server, to Anthropic’s API, which runs the Claude language model and returns output back through the same path to your browser.
That’s the whole data flow. It’s the same architecture as most AI tools in the legal space — including the ones your firm may have already approved. We’re not doing anything unusual. But we want to tell you specifically what that means.
What we ask you not to do
Don’t paste real client names, real matter details, real case facts, real financial figures, or real identifying information into any tool on this site.
The tools work with placeholders. [CLIENT], [MATTER], Client A, initials, pseudonyms, rounded or synthetic numbers — all of these produce output that’s just as useful. Every tool includes a one-click example that shows you what placeholder-mode input looks like.
You are a lawyer. You know your jurisdiction’s rules on disclosing client-identifiable information to third-party vendors. We’re not your ethics counsel. But we’d rather you use these tools safely with synthetic data than unsafely with real data, and we’ve designed them for the former.
What happens on our side
We do not log the content of tool inputs or outputs. Our server processes the request, forwards it to Anthropic, returns the response to your browser, and moves on. The only thing we record is a timestamp and the tool name — enough to know the tool is being used, not enough to know what you used it for.
Email addresses captured in the tool flow go to our email system for the nurture sequence. You can unsubscribe from any email. Unsubscribing deletes you from our list within seven days.
Intake form submissions (Full Practice Review, Casetext Migration, First Hire Readiness) are different from tool use. We retain those until 30 days after the report is delivered, then delete them. This is necessary to write the report. If you need earlier deletion, reply to the intake confirmation email and ask.
What happens on Anthropic’s side
Anthropic’s public policy is that API requests are not used to train their models by default. We’ve configured Prevaldi’s Anthropic API usage with Zero Data Retention (ZDR) enabled where available. Under ZDR, Anthropic processes the prompt, returns the output, and retains neither.
Anthropic’s full privacy commitments are at anthropic.com/legal. Their infrastructure is a real third party and we don’t control it — we’ve picked the most restrictive setting they offer.
What we will never do
We will not sell or share your data with advertisers. We have no advertisers. We have no analytics pixels from ad networks. We don’t have a Facebook or LinkedIn tracking pixel anywhere on this site.
We will not resell information about your practice to software vendors or recruiters. The research tools mentioned in the Casetext Migration Audit and the software stacks referenced in practice reviews are not shared with those companies. We don’t take referral fees.
We will not contact you outside the email flows you’ve opted into.
What you control
You can email us at [email protected] with any of the following requests and we’ll handle them within five business days:
- Delete my intake and report early (before the 30-day window)
- Remove me from the nurture email list
- Tell me exactly what information you have associated with my email
- Export everything you have on me
- Close my account, if I have one
What this document isn’t
This isn’t a contract. It’s a description. If you want a formal data processing agreement for your firm’s records, we’re a small company and we’ll sign the reasonable kind — reply to any email and ask.
If anything here changes materially, we’ll update this page and email anyone currently in our email system to tell them what changed and why.
Last updated: April 24, 2026
Questions? [email protected]