Enterprise
Content integrity for teams that publish
Teams that publish now have to answer the same question: which words came from a person, which from a model, and what record proves it. Enterprise answers it in three steps: detect, remediate, audit. It does not trade in the detector-evasion business. The record stays honest, or the tool does not run here.
Detect
Paste a draft or send one through the API and get back the same scored-tell report the free tool gives individuals, at volume.
The scorer is a fixed, documented rule set, not a model guessing at authorship. It names the pattern it hit and where, so a flag is a thing a human can read and act on.
Runs keep no words. The record stores counts, timestamps, and which rules fired, never your text.
Remediate
Rewrite the draft in a chosen voice with the same editor the free tool runs, in batch, at the limits of your plan.
The rewrite is provenance-preserving: it changes the writing, not the record. No stripping of C2PA credentials, no removing of watermarks, no claim that a model never touched the draft.
If a label is required on the output, the label stays. The edit takes the generic machine register out; it does not launder the draft.
Audit
Every run logs what changed and why: input word count, output word count, tells cut, the voice used, the account that ran it, and the timestamp.
Export the log as a report, the same downloadable report Pro includes, at whole-team scale.
The log is yours to keep under the DPA, and the 365-day retention applies to the counts and timestamps, not to drafts.
One law now sets the floor for machine-made text in the places your team publishes. Enterprise is built to keep the record that law asks for: labeled, marked, and traceable.
EU AI Act, Article 50
In force since 2 August 2026. AI-generated content must be labeled as such, and machine-readable marks are part of that regime. The audit log and the provenance-preserving rewrite give you the record a label points back to.
California AB 3211
Vetoed by Governor Newsom, so the bill never became law. There is no California watermarking requirement in force to comply with. For teams that publish across borders, the EU AI Act is the binding compliance driver right now.
We are not lawyers and this is not legal advice. The laws are named with their dates so you can check the obligations against your own counsel.
- 90 days at $299 a month.
- 3,000,000 words a month across the team.
- A DPA and onboarding included, with a named person on our side.
- You run the detect-remediate-audit loop on your own drafts before you commit to a year.
Goes to the contact form. Put the team size and monthly volume in the message and we will scope the pilot from there.
Questions
Is this a way to make AI writing pass as human?
No. The rewrite removes the generic machine register; it does not erase the history of the draft and it does not touch provenance metadata. If a label is required, the label stays.
What does the audit log store?
Counts, timestamps, which rules fired, and which account ran the edit. Draft text is kept for 0 seconds unless you donate a pair after a run, the same rule the free tier runs.
How is Enterprise different from Pro and Team?
Pro and Team are the self-serve plans for individuals and small teams. Enterprise adds SSO, org seats, the audit log, the usage API, the DPA, 365-day retention, onboarding, and an SLA.
Do you sign a DPA before the pilot?
Yes. The pilot price includes a DPA from day one. An audit feature needs a data agreement to mean anything.