DRAFT - version draft-2026-11-03. Not yet in force.

HideFirst Acceptable Use Policy

Effective date: [EFFECTIVE DATE]. This policy is part of the HideFirst Terms of Service (the "Terms"); defined terms have the same meaning. It applies to the Customer and every Authorised User.

1. Lawful and authorised use

1.1 Do not use the Services for anything unlawful, or to infringe anyone's rights.

1.2 Do not process data that you have no right to process, or in a way the law forbids.

1.3 Do not try to re-identify a person from pseudonymised output or from a Payload of a third party, or from any Payload by analysis rather than through your own alias book, and do not use anyone else's alias book. Restoring your own text with your own alias book is the purpose of the Services and is permitted.

2. Security and integrity

2.1 Do not attack, probe, scan or load-test the Services, or try to gain access to systems, accounts or data that are not yours, without our prior written permission, except good-faith security research under clause 2.3.

2.2 Do not interfere with the Services or with other customers' use of them, and do not introduce malware.

2.3 Responsible disclosure is welcome. If you find a vulnerability, report it to [security@hidefirst.com] with enough detail to reproduce it. Do not access or keep more data than you need to show the problem, and give us a reasonable time to fix it before disclosing it. We will not take action against good-faith research that follows this paragraph.

3. Protection of our technology (short form; see clause 6 of the Terms)

3.1 Do not extract, distil or copy the models or Model Materials, or train or fine-tune any model on Outputs or the Payload format.

3.2 Do not reverse engineer, decompile, disassemble or deobfuscate the SDK, the Python helper, the Extension, the WASM or the model bundles, except where mandatory law expressly allows it and clause 6.3 of the Terms is followed.

3.3 Do not query the Services systematically to reconstruct the model, word lists, thresholds or tables, and do not scrape the Services.

3.4 Do not benchmark or evaluate the Services to build or improve a competing product, and do not publish benchmarks without our prior written consent.

3.5 Do not get around rate limits, quotas, key restrictions, the OPRF shielding, signature checks or any other technical protection.

4. Keys and accounts

4.1 Keep API keys and passwords secret. Do not share them outside your organisation, sell or lend access, or resell the Services.

4.2 Do not create accounts or keys to avoid limits or a suspension.

5. The browser extension

5.1 Use the Extension only on sites and in ways that the terms of those sites allow. Do not use it to break the terms of service of ChatGPT, Claude, Gemini or any other chat site, or to automate those sites in ways that they forbid.

5.2 Do not repackage or redistribute the Extension, or load a modified copy.

6. Enforcement and reporting

6.1 If we believe this policy has been breached, we may act under the Terms: a proportionate suspension (clause 13), including immediately where security or intellectual property is at stake, and termination (clause 14), with no cure period for breaches of the IP or acceptable-use clauses. We will tell you what we found and why, where we can.

6.2 A breach may also make you liable under clauses 11.4 and 12 of the Terms, which are not capped.

6.3 To report a breach or abuse by anyone, or a security issue, write to [security@hidefirst.com]. For other questions about this policy, write to [legal@hidefirst.com].

6.4 We may update this policy under clause 22 of the Terms.