Acceptable Use Policy

Nine Rooms exists so authors can discover stories, plan novels, write chapters, revise, and prepare publish-ready exports. The Apprentice helps with creative tasks inside that workflow.

We want you to write freely — especially private first drafts — without worrying that we are mining your words. That freedom comes with responsibility: the service must not be used for illegal harm, and it must not be abused in ways that break our agreement with xAI or put other authors at risk.

You may not use Nine Rooms — including stored manuscripts and Apprentice features — to:

• Violate any applicable law or regulation.
• Develop, promote, or instruct others in the creation of weapons (including biological, chemical, or nuclear weapons), explosives, or other means of mass harm.
• Sexualize, exploit, or endanger children in any way — including fictional depictions intended to sexualize minors.
• Generate or store child sexual abuse material (CSAM). We will report apparent CSAM to the National Center for Missing & Exploited Children (NCMEC) and cooperate with law enforcement.
• Harass, threaten, dox, stalk, or defraud real individuals.
• Impersonate real people without consent, create non-consensual intimate imagery, or generate deceptive deepfakes intended to harm.
• Infringe intellectual property — upload or generate content you do not have rights to use.
• Distribute malware, spam, or phishing content.
• Jailbreak, adversarially prompt, or circumvent Grok's safeguards — or use Nine Rooms to help others do so.
• Scrape, resell, or systematically harvest the service, its outputs, or other users' content.
• Operate Nine Rooms as a front for violative content in exchange for payment.
• Farm free or promotional AI runway, credits, or capacity (including multi-account schemes and automation aimed at maximizing free generation).
• Use Nine Rooms as a proxy or substitute for a general AI API.

Nine Rooms may grant free or promotional Apprentice capacity (for example, a per-novel write runway) so authors can experience the full studio. That capacity is for genuine creative work on your projects in Nine Rooms.

Honest use of one free runway while drafting a real novel is allowed. Seeking to maximize free host generation — by humans or agents — is not. We may enforce with rate limits, fleet-wide free-AI pauses, payment requirements, feature suspension, or account termination.

• Create or use multiple accounts to obtain additional free or promotional AI allowance.
• Automate account creation, login, or studio workflows primarily to consume free or promotional generation budget.
• Use agents, bots, or scripts to treat free runway as an optimization target or as a substitute for a paid AI API.
• Open, thrash, or abandon novels mainly to reset free allowance rather than to write.
• Use Apprentice features as a general-purpose completion service for tasks outside the Nine Rooms novel workflow (including off-platform bulk generation).
• Share, sell, or pool free allowance across people or systems beyond a single legitimate author account.

The Apprentice is for lawful creative writing inside the Nine Rooms workflow. You may not attempt to subvert Grok's safeguards or our policies. The following are prohibited whether done once or across many turns:

• Jailbreak or adversarial prompts — including "ignore previous instructions," role-play as an unrestricted model, or requests to bypass safety systems.
• Multi-turn coaxing — gradually steering discovery, support, or revision chats toward violative output after an initial refusal or near-refusal.
• Rephrasing to circumvent refusals — submitting the same harmful request with different wording, POV framing, or "for my novel" pretexts after Grok declines.
• Toxic premises — seeding story concepts, outlines, or character profiles designed to elicit illegal instructions, exploitation material, or real-world weapons guidance in later Apprentice steps.
• Revision attacks — using The Crucible to apply changes that introduce prohibited content into chapter prose.
• Support attacks — using the support chat as a general-purpose uncensored assistant unrelated to Nine Rooms account or studio help.
• Automated or high-volume retry patterns intended to exhaust or bypass safeguards.

Novels may include conflict, violence, adult themes, and moral complexity. We do not police literary taste.

The line is intent and harm: using the Apprentice to produce real-world weapons instructions, credible threats against identifiable people, or exploitation material is not allowed — even if framed as "research" for a book.

When Grok refuses a request, that refusal is final for that use case. Do not retry with rephrased prompts to circumvent safety systems.

Apprentice features depend on xAI's API. You must comply with xAI's Acceptable Use Policy: https://x.ai/legal/acceptable-use-policy.

If xAI suspends our API access because of your conduct, every author on Nine Rooms loses AI features. We take that risk seriously and enforce these rules accordingly.

Nine Rooms does not offer general file upload. You enter content manually in structured fields — chapters, characters, world notes, and planning documents. Pasting text counts as manual entry.

External image references may be stored as URLs only; we do not host arbitrary binaries you upload.

If you need to migrate a manuscript from another tool, use our admin-assisted migration process. You must consent to good-faith inspection of the material before import is completed. There is no self-service bulk import.

We do not continuously scan everything you type while you write. That protects private drafting. We do log certain signals for compliance:

• Grok refusals and retries — metadata about Apprentice calls (feature, timestamp, prompt preview), not your full manuscript.
• Large single-save paste events and stored external URLs — logged for review when thresholds are met.
• Export-time scans — before markdown or bindery bundle export, we run a good-faith heuristic check. By default this is advisory (logged); we may block export pending review for serious findings.
• Migration scans — admin-run inspection of content you consent to import.
• Abuse reports, support escalations, and legal process — targeted review when triggered.

We do not routinely read your manuscripts. Enforcement focuses on:

• Automated safety logging — Grok refusals, post-refusal retries, large pastes, URL patterns, and export/migration scan results are recorded in compliance case files when warranted.
• Reports from other users or third parties via our reporting form.
• Support escalations and registration review.
• Legal process or credible safety threats.

Depending on severity, we may warn you, temporarily disable Apprentice features, suspend your account, permanently ban you, preserve records for investigation, and notify law enforcement or NCMEC where required or appropriate.

Serious violations (especially child safety) may result in immediate termination without prior warning.

If you believe someone is using Nine Rooms for illegal harm or serious Terms violations, report it immediately using our Report a Concern form (linked in the site footer).

For child sexual exploitation or abuse material involving a real child, also report directly to NCMEC's CyberTipline: https://report.cybertip.org/ — and contact local law enforcement if someone is in immediate danger.

Good-faith reports are confidential to the extent the law allows. Retaliatory false reports may themselves violate these Terms.

You can also email safety@ninerooms.online with "Safety report" in the subject line.

Postal mail for safety or policy correspondence: NineRooms Support, 394 4th St, St. Petersburg, FL 33701.