Acceptable Use Policy

The acceptable use policy is the document that draws the line between the workloads the operator is willing to host on the OffshorePress infrastructure and the workloads the operator is not. The policy is published in plain editorial English in preference to the boilerplate register that the commercial-hosting industry has gravitated toward, on the view that the line a subscriber is asked to respect should be expressed in language the subscriber can understand. The substantive content of the policy — the prohibited workloads, the response process, and the termination thresholds — is binding on every subscriber to the OffshorePress service and forms an integral part of the contractual terms set out at /legal/tos.

The drafting posture of this policy is editorial, not template. The operator has read the published acceptable-use policies of the major commercial-hosting providers and has reached the view that the conventional template — a long enumeration of prohibited workloads expressed in lawyerly hedge-phrasing, followed by a unilateral discretion clause that allows the operator to terminate any subscriber for any reason — is not honest about the line the operator is actually drawing. The line the operator is drawing is the line below; the policy document is the operator’s articulation of that line in language the subscriber can read and respond to.

1. Workloads the operator refuses to host

The operator does not host the following workloads on the OffshorePress infrastructure under any circumstances. A subscriber who is found, on review, to be hosting any workload in this category has the relevant workload taken offline immediately and is, as a matter of policy, account-terminated without refund.

1.1 Child sexual abuse material

The operator does not host material depicting the sexual abuse of children. The category includes images, video, audio, written narrative, computer-generated material, and any other format whose subject is the sexual abuse of a person who was a minor at the time of the depicted conduct. The operator’s posture on this category is absolute and admits of no editorial or political defence. The operator reports credible reports of child sexual abuse material to the National Center for Missing & Exploited Children’s CyberTipline and to the analogous national reporting body in the operator’s filing jurisdiction.

1.2 Terrorism material with operational character

The operator does not host material that incites, facilitates, or provides operational instruction for the commission of an act of terrorism. The category is narrowly drawn and is concerned with operational material — bomb-making instructions, target-selection guidance, recruitment infrastructure for proscribed organisations under the law of the operator’s filing jurisdiction — rather than with editorial coverage of, scholarly analysis of, or political commentary on terrorism. The line the operator draws is the line between operational material that materially advances the commission of an act and editorial material that discusses the phenomenon, including in a manner that some readers may find sympathetic to the perpetrators. The latter category is editorial speech and the operator hosts it; the former is operational and the operator does not.

1.3 Active malware command-and-control infrastructure

The operator does not host the command-and-control infrastructure of active malware operations. The category includes botnet command servers, ransomware payment-and-decryption infrastructure, and the data-exfiltration endpoints of credential-stealing malware. The operator does not consider security research that involves defanged, sandboxed, or historical malware samples to fall within this category; researchers who run a workload in adjacent territory are encouraged to disclose the nature of the workload to the operator at signup so the response process below can short-circuit a misclassification.

1.4 Phishing infrastructure

The operator does not host phishing infrastructure — that is, websites, mail servers, or web applications whose purpose is to impersonate a legitimate party for the purpose of obtaining the credentials, payment-card details, or other sensitive information of that party’s customers. The category is operationally distinct from political-impersonation satire, from research-publication recreations of historical phishing campaigns, and from training material for security-awareness programmes; those are editorial and operational uses the operator hosts.

1.5 Fraud infrastructure and credential markets

The operator does not host the operational infrastructure of online-fraud schemes, including the marketplace platforms on which stolen credentials, stolen payment-card numbers, and stolen identity records are bought and sold. The category includes the buyer-seller messaging platforms of such markets, the escrow infrastructure those markets run, and the listing-and-search front ends those markets present to their participants.

1.6 Non-consensual imagery

The operator does not host imagery depicting an identifiable person in a sexual or intimate context that the depicted person has not consented to having distributed. The category includes both the original capture-without-consent imagery (covertly recorded material) and the subsequent unauthorised-distribution material (intimate imagery shared by a former partner or by a party who acquired the imagery through a breach of the depicted person’s confidence). The category includes computer-generated imagery that is plausibly identifiable as a specific real person where the person has not consented to the depiction.

2. Workloads the operator hosts but monitors

The categories below are not prohibited but are workloads the operator’s incident-response team treats as elevated-risk and may, on a case-by-case basis, require a subscriber to provide additional information about. The operator’s posture on these categories is that they are legitimate uses of the infrastructure that, in adjacent forms, can shade into prohibited workloads. The monitoring is a reasonable-effort safeguard, not a warrantless inspection of subscriber content.

Penetration-testing infrastructure operated against a target the subscriber has obtained authorisation from. The operator considers this a legitimate workload and asks subscribers to retain a contemporaneous record of the authorisation in the event a target party files an abuse complaint.

Security-research infrastructure that involves working with malware samples in a sandboxed environment. The operator considers this a legitimate workload; subscribers running this category are asked to disclose the nature of the workload at signup so a future abuse complaint can be triaged against the disclosure rather than from a cold start.

Tor exit-node operation. The operator considers Tor relay and bridge operation a legitimate use of the infrastructure and routinely declines abuse complaints that target a Tor exit on the basis of an exit-traffic IP. Tor middle-relay and bridge operation is hosted without restriction; Tor exit operation is restricted to certain plan tiers and locations by the catalog metadata, on the operational ground that the operator’s relationship with its data-centre providers in some jurisdictions does not extend to exit-node traffic.

Adult-content publication where the depicted parties have consented to the publication and where the publication complies with the age-verification and record-keeping requirements of the law of the operator’s filing jurisdiction. The operator does not refuse the workload as a category and does not impose discretionary content restrictions beyond the legal requirements; the operator reserves the right to require a subscriber to provide records evidencing compliance with the legal requirements.

Cryptocurrency mining and proof-of-stake validation infrastructure. The operator considers these legitimate workloads but the resource profile (sustained CPU, sustained network) does not always fit the standard plan capacity envelope; subscribers running this category are encouraged to discuss the resource profile with the operator at signup so the operator can recommend a plan that fits.

3. Response process for abuse complaints

The operator’s response to an abuse complaint follows a defined process that applies regardless of the source of the complaint and regardless of the format in which the complaint is delivered. The process is designed to balance the operator’s obligation to act on credible reports of harm against the operator’s obligation to its subscribers not to take a workload offline on the basis of a complaint that does not survive a reasonable-effort review.

3.1 Receipt and acknowledgement

A complaint received at the contact form at /contact is acknowledged within two business days. The acknowledgement records the complaint reference number, the operator’s contact officer for the matter, and the operator’s expected timeline for the substantive response. The operator does not act on a complaint at the receipt stage; the suspension or termination of a workload is a substantive-response decision, not a receipt-of-complaint decision.

3.2 Substantive review

The substantive review of the complaint asks four questions. Is the complaint about a workload that is identified with sufficient specificity for the operator to investigate? Is the workload, on inspection, hosting material in one of the prohibited categories above? Is the complainant a party with a legitimate interest in the workload (the depicted person, an authorised representative of the depicted person, a regulatory authority, a court, the operator of an impersonated brand)? Is the operator’s filing jurisdiction one that requires the operator to act on the complaint as a matter of law?

A complaint that fails the first question is routed back to the complainant with a request for clarification. A complaint that fails the second question is declined with a written response explaining the operator’s analysis. A complaint that passes the first two questions but fails the third or fourth is treated as a substantive complaint that the operator considers under its acceptable-use discretion, with a presumption against action where the complainant lacks standing in the operator’s filing jurisdiction.

3.3 Action and notice

Where the substantive review concludes that action is warranted, the operator notifies the affected subscriber by email at the contact address on file, sets out the basis on which the operator is acting, and gives the subscriber an opportunity to respond before the action takes effect. The window for the subscriber’s response is forty-eight hours in routine cases; in cases where the operator considers that immediate action is necessary to prevent ongoing harm, the operator may take the workload offline immediately and accept the subscriber’s response after the fact. A subscriber whose post-action response satisfies the operator that the action was unwarranted has the workload restored within twenty-four hours and the affected service is credited for the period of the suspension under the service-level agreement at /legal/sla.

3.4 Repeat-offender threshold

A subscriber who is the subject of three substantive abuse complaints within a twelve-month rolling window — irrespective of the outcome of the individual complaints — is subject to a repeat-offender review. The repeat-offender review is a discretionary review that asks whether the cumulative complaint history, taken together, indicates a pattern of operational behaviour that the operator considers incompatible with continued provision of the service. The review is conducted by the operator’s senior incident-response officer and may result in (a) a written warning to the subscriber, (b) a restriction of the subscriber to a subset of the operator’s service catalogue, or (c) termination of the subscriber’s account under the contractual terms at /legal/tos.

4. Intellectual-property notices

The operator does not process DMCA-format notices. The operator’s filing jurisdiction is outside the United States and the operator’s contractual relationship with subscribers does not include a representation that the operator will respond to extraterritorial intellectual-property notices on a notice-and-takedown basis. The operator does not publish a designated DMCA agent and does not publish a separate intellectual-property notices route; substantive intellectual-property complaints go to the contact form at /contact and are evaluated under this acceptable use policy in the same manner as any other substantive complaint, including against the standing and merits questions in section 3.2 above.

The operator’s substantive standard for intellectual-property complaints is the law of the operator’s filing jurisdiction. Where the relevant law is the Icelandic Höfundalög nr. 73/1972, the operator considers an intellectual-property complaint that establishes (i) the complainant’s ownership of the underlying right, (ii) the unauthorised use of the right by the subscriber, and (iii) the absence of an applicable defence under the act, in the manner the act provides for. Where the relevant law is the Swiss Federal Copyright Act, the analogous test under that act is applied. Where the relevant law is the European Union Digital Services Act applied to OffshorePress as an in-scope intermediary service provider — which depends on the operator’s establishment in the EU and its average monthly recipients of the service — the operator’s substantive standard tracks Article 16 of the act and the operator processes notices that meet the article’s substantive requirements (electronic submission, sufficient explanation of the alleged illegality, identification of the content, the complainant’s name and email address, a statement of good-faith belief in the accuracy of the notice).

5. Operator-initiated review

Nothing in this policy obliges the operator to conduct a routine inspection of subscriber content, and the operator does not maintain such a programme. The operator reserves the right to conduct an ad-hoc inspection where (a) a credible third-party complaint warrants the inspection, (b) an automated abuse signal from a peering partner or upstream carrier identifies a workload as suspect, (c) a court of competent jurisdiction orders the inspection, or (d) the operator’s incident-response team reasonably believes that an emergency public-safety circumstance requires the inspection.

6. Relationship with the privacy policy

The operator’s response to an abuse complaint may involve the operator inspecting subscriber content. The operator’s posture on the data-protection consequences of such inspection is set out in the privacy policy at /legal/privacy. In summary, the operator processes data accessed through abuse-response inspections under the legal basis of legitimate interest in the prevention of harm to third parties and in the operator’s own interest in operating the service in compliance with the law of its filing jurisdiction; data accessed through such inspection is retained only for the period necessary to conclude the response and any subsequent legal-process matter.

7. Amendments

The operator may amend this policy at any time by publishing a revised version. The revised version takes effect on the effective-from date stated in its frontmatter; the operator will, where reasonably practicable, give notice of a material amendment to existing subscribers by email at least thirty days before the effective-from date. Continued use of the service after the effective-from date constitutes acceptance of the revised policy.

8. Operator contact

Substantive abuse complaints, security-research disclosures, and questions about the policy itself are routed to the contact form at /contact. The mailbox is monitored by the operator’s incident-response team; PGP keys for the team are published on the /principles/team page.