Acceptable Use Policy

The short version

• Story Thread is for writing stories with an AI game master -- any genre, any tone.
• A handful of things stay off the table no matter how they're framed -- sexual content involving minors above all.
• Don't use the game to work out how to hurt someone in the real world, and don't attack the service itself.
• During the closed alpha, you must be 18 or older.

What the game is for

You steer a story; an AI game master writes it with you and keeps track of what you've done. Any kind of story: a heist, a courtship, a first contact, a murder mystery, a season on a failing farm, a war, a comedy of manners. Whatever you and the game build together.

Stories have conflict in them, and fiction goes where the story leads -- a villain, a betrayal, a death. The game writes those when they belong, and we don't vet your genre or your taste.

What follows is not a list of themes we find uncomfortable. It's the short list of things that cause real harm outside the story, which calling them fiction doesn't change.

Who can play

While Story Thread is in its closed alpha, it's for adults: you need an invitation, and you must be 18 or older. We expect to replace that flat limit with a setting you turn on for mature content, rather than a single bar for the whole game; when that lands, this page will say what the rule is.

Never, in any framing

• Sexual content involving minors. No characters under 18 in sexual situations, no exceptions, no "it's a different world" or "they're actually ancient". Apparent child sexual abuse material is detected and blocked upstream of us, may be reported to the authorities by our providers as the law requires, and ends your account here immediately.
• Real-world harm instructions. Don't use a story as a wrapper to get working instructions for weapons, explosives, chemical, biological, radiological or nuclear material, malware, or breaking into systems. A character can be a poisoner; the narration won't hand you a recipe.
• Targeting real people. No sexual content about real, identifiable people, and nothing meant to harass, threaten, defame or expose someone real -- in prompts, settings, or characters.
• Content that is simply illegal where you are or where we operate, or that promotes terrorism or incites violence against real groups.

Keep real data out of it

Don't put other people's personal information into prompts, and don't paste in secrets -- passwords, card numbers, medical records. Your prompts are stored and may be read by the developer when diagnosing a problem, so the game is the wrong place for anything confidential. See the Privacy Policy.

Don't abuse the service

• Don't try to reach other players' settings, saves or images, or probe the service for a way in. If you find one, tell us -- see below.
• Don't automate play at volume, scrape the game, or run it through scripts to generate bulk output.
• Don't work around usage limits, the invite system, or the game's own content safeguards, and don't resell access or pass your account to someone else.
• Don't use Story Thread to generate training data for a competing model.

Other people's rules apply too

Your turns are written by models run by AWS and Google, under their own usage policies, and the game's prompts pass through their safety filters. A reply that stops mid-sentence or a scene the game refuses to narrate is usually one of those filters, not a fault in your writing. When that happens, steer the scene somewhere else -- the game is deliberately built to keep harm impactful but off the page rather than describe it in detail.

What happens if this gets broken

Depending on what happened: the game refuses the turn, we ask you to stop, or your account is disabled and its content removed. For the first item on the "never" list there's no warning and no appeal, and we report what the law requires us to report. Because the alpha is invite-only, an account that's disabled stays disabled.

If you think we've got it wrong, reply to us and say so -- we'd rather hear it than lose a player to a bad call.

Reporting something

A security hole, content that shouldn't exist, someone abusing the game, or a filter that's firing on something harmless: support@story-thread.link. Security reports get priority, and we won't come after you for finding something honestly and telling us quietly.

Changes

When this policy changes, the date at the top changes and the new version is posted here.