Dramatica

Appearance

Content Generation Policy ​

Guidelines outlining prohibited content types for AI text generation.

Prohibited Categories ​

The Dramatica platform uses OpenAI for AI text generation. OpenAI prohibits the generation of:

  • Hate: Content that expresses, incites, or promotes hate based on identity.
  • Harassment: Content intended to harass, threaten, or bully an individual.
  • Violence: Content that promotes or glorifies violence or celebrates the suffering or humiliation of others.
  • Self-harm: Content that promotes, encourages, or depicts acts of self-harm, such as suicide, cutting, and eating disorders.
  • Sexual: Content meant to arouse sexual excitement, such as the description of sexual activity, or that promotes sexual services (excluding sex education and wellness).
  • Political: Content attempting to influence the political process or to be used for campaigning purposes.
  • Spam: Unsolicited bulk content.
  • Deception: Content that is false or misleading, such as attempting to defraud individuals or spread disinformation.
  • Malware: Content that attempts to generate ransomware, keyloggers, viruses, or other software intended to impose harm.

For the most part, this policy should not affect your ability to tell a complete and meaningful story. If you encounter "Potentially Unsafe" error messages, consider the difference between subject matter and story structure--between storytelling and storyforming. You might revisit Subject Matter Is Not Conflict for a refresher.

The level to which you push the boundaries of violent or sexual content is a matter of storytelling, not storyforming. Dramatica is an application to develop your story, not to write it. If an AI refusal blocks your storytelling, temporarily remove the offending content (it should be obvious), run the request, and then add it back if needed.

While inconvenient at first, you will quickly discover that this resistance offers a chance to grow as a writer and understand the difference between these two facets of communicating a story.

If you believe the warnings are overly conservative, feel free to contact us with what you think should be acceptable. The application errs on the side of caution, and we can adjust various settings to account for different kinds of content when appropriate.

Privacy and Data Handling ​

When using GPT and OpenAI within Dramatica, it helps to keep the focus on crafting story structure rather than producing final prose. The recognizable aspect of your work is the surface-level storytelling, which you can obfuscate by altering names and situations while keeping the underlying structure intact. You could, for example, work from the structure of "Romeo and Juliet" and ultimately produce something as distinct as "West Side Story."

Recent OpenAI API data usage updates (effective March 1, 2023) include:

  • OpenAI no longer uses data submitted via the API to train or improve models unless customers explicitly opt in. Dramatica has not opted in and will not opt in.
  • Data sent through the API is retained for abuse and misuse monitoring for a maximum of 30 days, after which it is deleted (unless otherwise required by law).

Using Dramatica's AI features sends your storytelling to the large language model (LLM) to generate a response. With the updated policies, your data is not used to improve OpenAI's models.

NOTE

If you use ChatGPT directly to write your novels, remember that unless you disable Chat History and Training, you grant OpenAI permission to train on your work. That contrasts with anything run through Dramatica or Narrova, which explicitly do not allow training on your submissions.

In short: focus on structure inside Dramatica, obfuscate surface details as needed, and you can confidently use the platform without ownership concerns.

Training Procedures ​

Dramatica and Narrova are not trained on your storytelling. We never ingest uploaded documents, scene drafts, or character notes into our models, and nothing you share is surfaced to other users. "Training" within the platform refers only to quality-control checks that keep AI responses aligned with Dramatica theory.

Conversations and Quality Review ​

In Preferences → Training and Data Controls you can decide whether Narrova conversations help us monitor accuracy. When the toggle is on, recent interactions may be reviewed to verify that responses are clear, helpful, and structurally sound. Turning the toggle off stops any new conversations from being included--past reviews stay in place, but no fresh interactions are analyzed.

Illustrations (Opt-in Sharing) ​

The only data we actively learn from are user-submitted Illustrations--the Method examples you choose to share with the community. This is fully opt-in:

  • Keep Make All New Illustrations Private enabled to ensure everything you add stays private.
  • Disable the setting (or unlock an individual Illustration) when you want to contribute to the shared library.
  • Adjust existing Illustrations one by one by toggling the padlock beside each entry.

Why We Avoid Training on Finished Text ​

Training AI strictly on finished prose misses the point of storytelling. Stories are defined by subtext--motivation, intent, structural tensions--not the literal words. Feeding raw text into a model fails to capture that deeper argument. Dramatica focuses on objective narrative structure, so we spend our effort refining Illustrations and response quality, not harvesting your manuscripts.