Skip to content
Solo Pros & Small Business · Client Confidentiality

Use ChatGPT and Claude on Your Client Work — Without Breaking Your NDAs

PortEden replaces client identifiers, contract detail, and confidential material with placeholders before your prompt reaches OpenAI or Anthropic. Use any AI on real work; the model never sees the underlying client.

See pricing

Free for solo pros · No credit card · Per-client audit log built in

Maps to
GDPR
CCPA
SOC 2
EU AI Act
HIPAA
The Risk

What Goes Wrong Without PortEden

You Paste a Client Brief Into ChatGPT for a First Draft

That brief carries the client's confidential plans, pricing, and customer detail — every field your NDA explicitly covers. Sending it to OpenAI without sub-processor authorisation is the kind of disclosure that makes a renewal conversation go sideways the moment a procurement audit catches it.

You Use Claude to Edit a Contract or Statement of Work

The document holds the client's name, the deal terms, and sometimes a counterparty's confidential schedule. Claude's context window now holds the cleanest copy of the agreement, retained per Anthropic's policy, not the contract you actually signed.

You Connect Outlook or Gmail to an AI Assistant for Inbox Triage

Subject lines like "Acme — confidential roadmap, Q3 launch" and the strategy thread itself flow into the AI. One prompt-injection or subprocessor change and a client's confidential material lives somewhere you can't reach to delete it.

How PortEden Protects Every Client

Client-Confidential Material, Redacted Before It Reaches the Model.

PortEden inspects every field your AI is about to see. Client and counterparty names, contract terms, customer PII, and free-text confidential material are replaced with placeholders at the boundary — never sent to OpenAI or Anthropic.

Your data
PortEdenRedact
Your AI
Claude
ChatGPT
Copilot
Gemini
Grok
Safe
Sensitive
Redacted
Compliance Reality

What Your NDAs and Privacy Laws Actually Require When You Use ChatGPT or Claude on Client Work

  • Every NDA you've signed restricts disclosure of client confidential information to authorised parties for the contracted purpose. Pasting that information into a third-party AI without authorisation is exactly the disclosure the contract forbids — and the indemnity letter will land back on you.
  • GDPR Article 28 makes you a processor of every EU client's data and CCPA's service-provider rules cap how that data can be used. Most consumer ChatGPT and Claude tiers do not sit inside either contract chain — the moment your prompt contains client personal data, you're outside it.
  • OpenAI and Anthropic log prompts on consumer tiers for abuse monitoring and may train on free-tier conversations. "I disabled training" is not the same as "the prompt was never logged" — your client's GC will read the privacy policy more carefully than your settings page.
  • Procurement and security reviews now ask solo vendors for an AI tool inventory, a sub-processor list, and per-client data-flow maps. Saying "I sometimes use ChatGPT" is the answer that loses you the renewal or the next deal.
GDPR Article 28 — Processor
The Solution

Built For Professionals

Client-Confidential Redaction in Gmail and Outlook

Replaces 50+ identifier types — client and counterparty names, contract terms, customer PII, financial detail, and free-text confidential material — with placeholders in under 200 ms before the prompt leaves your perimeter. The AI receives only the redacted version; client material never reaches OpenAI or Anthropic.

Per-Client Profiles With Strict Isolation

Each client account gets its own redaction profile and its own audit log. Confidential material from your largest client never crosses into a prompt for another client, and a procurement reviewer can see the per-client policy without needing access to your raw inbox.

Procurement-Ready Audit Log

Per-client, per-prompt record of every AI use, every redaction outcome, and every sub-processor it touched. Drops directly into the AI tool inventory, sub-processor list, and data-flow map a Fortune-500 procurement team now expects from solo vendors.

Works With Claude, ChatGPT, and Copilot — No Workflow Change

Keep using Notion, Google Docs, Slack, Asana, Linear, or whatever your one-person stack is built on. PortEden sits in front of any AI tool — no plugin, no per-client browser extension, no training your AI assistant on a new interface. Five-minute setup.

Works With What You Already Use
Gmail
Gmail
Inbox-side redaction for AI replies on client email
Outlook
Outlook
Microsoft 365 inbox-side redaction
Google Calendar
Google Calendar
Strip client names from meeting and review titles
Slack
Slack
Strip client identifiers from internal AI-summary threads
Notion
Notion
Per-client redaction across your wiki and project docs
With and Without PortEden

The Same Workflow, Two Very Different Outcomes

Drafting a Client Memo or Proposal With ChatGPT
Without
Client name, confidential plans, and pricing sent to OpenAI in plain text — an NDA disclosure with no sub-processor authorisation in the chain.
With
Client identifiers and confidential material replaced with placeholders before the request leaves your network. The AI drafts the memo; PortEden re-hydrates the client identity locally.
Editing a Contract or SOW With Claude
Without
Counterparty names, deal terms, and confidential schedules sent to Anthropic. Retained per their policy, not the agreement you signed with the client.
With
Document structure and language reach the model; identifiers, deal terms, and counterparty names are placeholders. The AI suggests edits without seeing whose contract.
Inbox Search With Copilot or Gemini
Without
Every matching email — client briefs, attached SOWs, customer lists — sent to the AI in plain text. Cross-client leakage if the search spans accounts.
With
Email content reaches the model with each client's confidential material redacted under that client's profile. The AI ranks results without crossing the client boundary.
Procurement or Security Review From Your Largest Client
Without
You scramble to assemble an AI tool list, sub-processor map, and a data-flow trail from screenshots and inbox archeology. The renewal slips while the GC waits.
With
Per-client audit log of every AI use, every sub-processor, and every data flow — exportable as the procurement pack the GC actually wanted to see.
Adding a Contractor or Junior to Your Workflow
Without
You either trust them with raw client material in the AI tools they already use, or you don't add them. There's no third option that scales.
With
Contractor inherits your per-client profiles automatically. They see redacted prompts, you see the audit log of what they did, and the client's NDA stays intact.
Try It on Your Largest Client

Five-Minute Setup. Free for Solo Pros.

Connect Gmail or Outlook via OAuth. Spin up a per-client profile. Keep using ChatGPT or Claude exactly the way you do today — with each client's confidential material protected by default.

See pricing

Frequently Asked Questions

I'm a solo consultant — do I really need this, or is the free ChatGPT tier fine?
If you've signed any NDA, MSA, or DPA with a client, the free tier almost certainly is not fine. Most NDAs restrict disclosure to authorised parties for the contracted purpose; consumer ChatGPT is neither. PortEden's free tier closes that gap without changing how you use the AI — the model still sees the structure of the work, just not the client identifiers and confidential material.
My vertical isn't on your for-professionals menu — does this still cover me?
Yes. The Professionals page is the catch-all for solo consultants, coaches, designers, writers, freelance engineers, fractional execs, executive assistants, and anyone running a one-person business that touches client material. The vertical-specific pages (lawyers, CPAs, therapists, etc.) just add the regulator-specific copy on top of the same redaction engine you'd use here.
How does PortEden help when a client asks for a security review?
The audit log exports as a per-client AI tool inventory, sub-processor list, and data-flow map. Those are the artifacts the SIG, CAIQ, and TRM-style questionnaires now ask solo vendors for. We've shipped templates that drop straight into a typical Fortune-500 procurement packet — turning a panic project into a 30-minute task.
Will PortEden change my workflow or my AI tools?
No. PortEden runs in front of your AI, not your tools. You keep using Claude, ChatGPT, Copilot, Gemini, Notion, Slack, Asana, Linear, Google Docs, or whatever your one-person stack is built on. The redaction layer sits between any AI tool you launch and the data that AI is about to read — same shortcuts, same UI, same models.
What happens when a client's regulated content (PHI, NPI, return data) flows through?
Per-client profiles can ride stricter rules per account. A healthcare client triggers PHI redaction; protected fields never reach the AI vendor. A financial client triggers Reg S-P / GLBA-aware controls. If your work spans regulated and unregulated clients, the policy follows the account — you don't have to remember which rule applies to which inbox thread.
Can a contractor or junior I add to my practice use it without exposing my clients?
Yes. Add the contractor to a per-client profile and they inherit your redaction defaults automatically. They see the same redacted prompts you do, you see the audit log of what they did and which sub-processor saw what, and the client's NDA stays intact through the whole chain.
What does it cost and how long does setup take?
There's a free tier for solo evaluation. Individual pros are $19/mo; small businesses are $49/mo per seat for 2–10 seats. Setup is under 5 minutes on Gmail or Outlook + ChatGPT or Claude. Multi-client professional services typically take a half-hour to spin up per-client profiles for the top accounts.

Keep Exploring

Ready to Use AI on Client Work Without Breaking Your NDAs?

Five-minute setup. Free for solo evaluation. Per-client audit log and procurement-ready AI sub-processor map from day one.

See pricing

Vertical-specific (lawyer, CPA, therapist…)? Pick your industry →