Skip to content
Legal

Privacy Policy

Last updated: September 24, 2026

Local Processing

Local device processing and optional hosted features have separate data flows

Review Before Sharing

Hosted tools receive the content you choose to submit

Your Data, Your Control

Choose what you submit to hosted tools; review each feature's data notice

1. Introduction

THOX.ai LLC. ("THOX.ai," "we," "us," or "our") is committed to protecting your privacy. This Privacy Policy explains how we collect, use, disclose, and safeguard your information when you use our edge AI device, website, and related services (collectively, the "Services").

By using our Services, you agree to the collection and use of information in accordance with this policy. THOX.ai™ and the THOX.ai logo are trademarks of THOX.ai LLC.

2. Information We Collect

2.1 Device Usage Data

When you use a local device workflow without enabling a hosted connection, that workflow does not send us:

  • Your source code or prompts
  • AI model inputs or outputs
  • File contents or project data
  • Any data processed by the AI models

2.2 Optional Analytics

With your explicit consent, we may collect anonymized usage statistics to improve our products:

  • Device health metrics (temperature, memory usage)
  • Firmware version information
  • Aggregate usage patterns (no personal data)
  • Error logs for troubleshooting

2.3 Website and Account Data

When you use our website or create an account, we may collect:

  • Name and email address
  • Billing and shipping information
  • Communication preferences
  • Support ticket contents
  • Content you submit to hosted tools, as described in the feature notice before you use them

2.4 Edge Lab hosted evaluation

Edge Lab verifies your THOX account and confirmed email before an evaluation. Submitting an evaluation sends your request text and selected options to this website. Only requests marked public and submitted with your explicit inference consent are forwarded to the selected hosted AI service shown in the demo. The service may be Vercel AI Gateway or the THOX inference gateway at llm.thox.ai. When Vercel AI Gateway is selected, the model-serving route is restricted to Google Vertex AI: Vercel and Google Vertex AI receive the content needed to process the request. Review the gateway, model and provider disclosure before consenting. This is hosted processing, not on-device inference. Use only public or synthetic examples; do not submit personal, sensitive or confidential information. Selecting the public label does not automatically detect or remove sensitive content.

The Edge Lab website adapter does not persist request text or generated receipts, and does not log provider responses or credentials. It stores account-identifier quota buckets, usage counts and expiration times to enforce shared and per-account limits. These quota records do not contain prompts or generated outputs. Expired counters are removed on a later quota request, rather than immediately when they expire. Hosting, authentication and gateway services have their own operating and retention policies; Edge Lab does not promise zero retention across those services.

Results are proposals for human review. Edge Lab does not execute actions or connect to your devices. The separate browser sample uses deterministic rules and does not send its entered text to an AI service. Marketing analytics scripts are excluded from the Edge Lab page. See Edge Lab for availability and the request consent notice.

2.5 ThoxEmployee hosted demo

The hosted demo stores the employee profiles, task briefs, generated text and review history you create in a private guest workspace. A necessary session cookie grants access to that workspace. Its value is stored as a digest on the server. Access expires after 24 hours; an hourly cleanup removes expired workspace content. Quota records use a keyed digest of your network address and a session digest, rather than storing the raw address in the demo tables. They cover the current and previous two UTC dates; older records are removed on the next request or hourly cleanup.

Before a live run, you must acknowledge that the task title and brief, selected employee names, roles and instructions, and earlier written contributions will be sent through Vercel AI Gateway to OpenAI. Do not submit sensitive or confidential information. You can export your work or clear the guest workspace. Clearing does not reset usage limits, erase operational infrastructure logs, or retract content already sent to a provider. See the demo notices for details.

3. How We Use Your Information

We use the information we collect to:

  • Process and fulfill your orders
  • Provide customer support
  • Send important product updates and security notices
  • Improve our products and services (with anonymized data only)
  • Comply with legal obligations
  • Run the hosted workflows you request and enforce usage and abuse limits

4. Data Sharing and Disclosure

We do not sell your personal information. We may share data with:

  • Service providers: Payment processors, shipping carriers, and cloud infrastructure providers who help us operate our business
  • Optional hosted AI: The provider named in a feature's notice receives the inputs needed for a live request after your explicit action
  • Legal requirements: When required by law or to protect our rights
  • Business transfers: In connection with a merger, acquisition, or sale of assets

5. AI Model Providers

Data handling depends on the model and connection you choose. Local device processing is separate from website demos and other hosted connections.

  • Local workflows: Inputs stay on the device when the workflow uses local inference and no hosted connection is enabled.
  • Hosted features: Read the feature notice for its provider, submitted content and storage behavior before running it.
  • Edge Lab: Consented live requests use the hosted gateway and model-serving provider disclosed in the demo. Provider processing and retention depend on that service; Edge Lab does not promise on-device processing, private-only routing or zero provider retention.
  • ThoxEmployee: Live text generation uses OpenAI through Vercel AI Gateway. Processing and retention follow the applicable Vercel AI Gateway data controls and OpenAI API data controls; the demo does not promise zero provider retention.

6. Data Security

Hosted demos use feature-specific controls:

  • HTTPS connections to the website and inference gateway
  • Server-only provider and database credentials, with restricted database access
  • Explicit consent and bounded inputs before live inference, with separate usage limits for each demo
  • Edge Lab: verified-account access and validation of generated proposals
  • ThoxEmployee: visitor-scoped state, revision checks, mandatory human review and expiry cleanup
  • These controls do not assert a compliance certification

7. Your Rights

Depending on your location, you may have the right to:

  • Access the personal data we hold about you
  • Request correction of inaccurate data
  • Request deletion of your data
  • Opt-out of marketing communications
  • Data portability

8. Children's Privacy

Our Services are not intended for children under 16. We do not knowingly collect personal information from children under 16. If you believe we have collected information from a child, please contact us immediately.

9. International Data Transfers

Your information may be transferred to and processed in the United States. We ensure appropriate safeguards are in place to protect your data in compliance with applicable laws.

10. Changes to This Policy

We may update this Privacy Policy from time to time. We will notify you of any material changes by posting the new policy on this page and updating the "Last updated" date.

Contact Us

If you have questions about this Privacy Policy or our data practices, please contact us at:

THOX.ai LLC.

Designed & Manufactured in the USA

privacy@thox.ai