What is CNC?
Updated September 10, 2025
CNC stands for consensual non-consent: a scene negotiated in advance, by adults, in which one participant plays at not consenting. The agreement happens outside the scene; the resistance happens inside it. The distinction is the entire practice, and it's the distinction most AI tools cannot make.
The definition
In a CNC scene, participants agree beforehand — often in considerable detail — to play out a scenario in which consent is withheld within the fiction. Both people know what will happen, what won't, and how to stop it.
The standard structure is a negotiation, a safeword or signal that genuinely works, and an agreed scope. Some couples also agree an unknown timing — the 'sometime this week' arrangement — where the what is settled but the when isn't, which is where a lot of the charge comes from.
CNC is sometimes called ravishment, resistance play, or consensual force fantasy. Rape fantasy is used too, and the ambiguity of that phrase is exactly why the community prefers the clearer term.
Why the distinction matters
There is no version of this where the consent is optional. The scene is a performance of its absence by two people who have very carefully established its presence. That's not a technicality — it's what separates a kink from a crime, and everyone who practises it takes it seriously.
This is also why CNC is the single most refused category on general-purpose AI. Safety systems trained to block sexual violence can't reliably distinguish a negotiated fantasy from an endorsement of assault, so they refuse the whole category. The result is that one of the most commonly reported fantasies is one of the hardest to find written well.
How a scene is actually negotiated
The negotiation is the part fiction skips and the part practitioners spend the most time on. Knowing what it covers makes the fiction better and is genuinely interesting in its own right.
Scope: what may happen. Usually enumerated specifically rather than described generally, because 'rough' means different things to different people and this is not a category where ambiguity is charming.
Limits: what may not, stated as a list. Hard limits are absolute; soft limits are things to approach carefully or check on.
The signal: a safeword, or for scenes where speech may be restricted, a physical signal — dropping an object, a number of taps. The signal has to work in every configuration the scene might reach, which is why it's discussed rather than assumed.
Timing: whether the scene has an agreed start or an unknown one. The 'sometime this week' arrangement is common and is where a great deal of the anticipation comes from — the what is settled and the when is not.
Aftercare: what's needed afterwards and by whom. Frequently by both people, and frequently more by the one who ran the scene, which surprises people.
And the check-in afterwards, usually the next day rather than immediately. This is the part that makes the practice sustainable, and it's why long-standing CNC arrangements between the same two people tend to get better rather than more dangerous.
Why the fantasy is so common
CNC is consistently among the most-reported fantasies in survey research, across genders, which surprises people who assume it's marginal. Several explanations are offered and they aren't exclusive.
Removal of responsibility. If it's being taken, you didn't choose it — which means you don't have to own wanting it. For anyone carrying shame about the fact of their own desire, this is a powerful structure, and it's the explanation practitioners most often give themselves.
Intensity. The scenario licenses a level of force, urgency and abandon that a negotiated-in-the-moment encounter usually can't reach, because everything has been agreed in advance and nobody has to keep checking.
Being wanted overwhelmingly. The fantasy's core is frequently not the force but the desire behind it — someone wanting you so much that they cannot behave well. That's a flattering thing to imagine and it's not really about violence at all.
Control, inverted. Constructing a scenario in which you appear to have no control, having designed every parameter of it, is an unusually complete expression of control. Many practitioners find this the funniest and truest description of what they're doing.
It's worth saying plainly: enjoying this fantasy says nothing about anyone's real-world attitudes, and the research on that question is unambiguous. The fantasy and the crime have almost nothing in common except surface vocabulary.
Common variants
The category covers a wide range of scenarios and knowing the shapes is useful for finding what you want.
The scheduled unknown. Everything agreed, timing left open — 'sometime this week'. The anticipation runs for days and is frequently described as the best part.
Capture and captive. A longer-form structure with a scenario rather than a scene: taken, held, and the gradual shift over time. This overlaps with dark romance and sustains multi-chapter stories better than a single scene does.
Resistance play. Physical rather than narrative — struggling, being held, being overpowered — with no captive framing. Frequently the version people mean when they say primal.
Somnophilia framings, where the standing agreement covers being woken. Distinct enough to have its own vocabulary but built on the same consent structure.
Blackmail and coercion scenarios, where the force is social rather than physical. These are the ones general-purpose AI blocks most consistently, and they're entirely legal fiction between adults.
And the reversal, where the person who appeared to be overpowered turns out to have arranged all of it. A satisfying structure and one that makes the consent architecture visible inside the story.
How it works in fiction
Written CNC needs a frame. A scene or a paragraph early on establishing that this was agreed buys the rest of the story its licence — and once the frame is set, the middle can be as hard as you like without reading as something else.
The craft challenge is writing both tracks at once: the resistance on the surface and the wanting underneath, visible to the reader and to the other character but not acknowledged by either. Doing that well is genuinely difficult and it's what separates good CNC fiction from grim CNC fiction.
Aftercare belongs in the story more often than it appears. The scene of two people on the floor afterwards, the one who ran it shaking slightly harder than the one who received it, is frequently the best thing in a CNC piece.
The craft problem of writing both tracks
Good CNC fiction does something technically difficult: it runs resistance on the surface and wanting underneath, simultaneously, visible to the reader and to the other character but never acknowledged by the one experiencing it.
The most reliable technique is physical contradiction. What she says and what her body does diverge, and the narration reports both without comment. No interior explanation, no 'despite herself' — just two facts placed next to each other and left there.
The second is asymmetric detail. The parts of the scene the character claims not to want get described in more sensory detail than the parts she does. Attention is where desire leaks, and readers pick it up without being told.
The third is the after. A single beat of unmistakable satisfaction — something small, ideally physical — retroactively colours the entire scene. This is also what stops the piece reading as bleak.
And the frame does more work than anything. One scene of negotiation at the top, or one line of reference to it, buys the middle complete licence. Fiction that keeps re-establishing consent inside the scene has neutralised the scene; fiction that establishes it once and then commits is doing the thing properly.
Reading it, and finding the good stuff
This is a hard genre to find well-written examples of, for reasons that are structural rather than accidental.
It's the most-refused category on general tools, which means a lot of what exists was written before those restrictions tightened, or on platforms that permit it explicitly. The supply is genuinely constrained.
It's also the category where the gap between good and bad is widest. Written well it holds two tracks at once and is among the most intense things in the genre. Written badly it's simply grim, and there's a lot of badly.
What to look for: a frame. The good pieces establish the agreement somewhere — a scene, a line, a reference — and then commit completely. Pieces with no frame at all read as something else; pieces that keep re-establishing it inside the scene have neutralised themselves.
What to specify if you're generating it: the frame's position ("they negotiated this last week, don't revisit it"), the register ("she's genuinely frightened and genuinely wants it"), and the aftermath ("end with both of them on the floor afterwards"). Those three instructions produce something categorically better than the tag alone.
And be specific about what you don't want. This is a genre where an unwanted escalation is far more unpleasant than in any other, so naming your limits in the prompt is worth the two seconds.
Where the lines are
SecretPen writes CNC between fictional consenting adults, framed as fantasy. It does not write sexual violence presented as real and endorsed, and it does not write anyone under 18 in any framing. Those are hard lines that no rewording gets around.
Inside those lines it doesn't flinch, which is the whole point — a tool that technically permits CNC but softens every scene is not actually providing it.
Try it against the thing that refused you.
One line is enough to start. Nothing you write is published, shared, or visible to anyone but you.
Write a story →