Does Adobe Firefly allow NSFW content? What the Generative AI User Guidelines say

Adobe does not allow explicit sexual content in its generative AI features. The Generative AI User Guidelines say users must not create explicit sexual material or explicit nudity, and the same list bars depicting nude minors or minors in a sexual manner. Adobe says it can disable accounts for any violation and that it may review prompts, inputs and results to filter content.

For adult game developers, the guidelines matter because Adobe’s tools sit inside widely used art software. Generative features in an image editor are used for backgrounds, extending canvases and cleaning up concept art, and the rules decide which of those jobs are safe to run on character art that is close to the line.

This page sets out what the guidelines say about content, review, third-party rights, Content Credentials and customised features, and what they mean for adult game art. It describes the guidelines as they stood on 4 November 2026, with the page showing a last updated date of 15 May 2026. The text refers to “Adobe’s generative AI features” and does not name Firefly or any other product, so whether a given product such as Firefly is covered should be confirmed in the product itself.

Key points

• Explicit sexual material and explicit nudity are barred. They head the list of things users must not create.
• Nude minors and sexualised minors are barred. The same list covers them.
• Prompts, inputs and results can be reviewed. Both automated and manual methods are named.
• Adobe can disable accounts. The right applies to any violation.
• Third-party copyright rules reach prompts and reference images.
• Content Credentials must stay. Removing or altering them is prohibited.

What the guidelines say about content

The Generative AI User Guidelines show a last updated date of 15 May 2026. They open with a short statement: they govern use of Adobe’s generative AI features, and Adobe reserves the right to disable accounts for any violation.

The first section is titled “Be Respectful and Safe”. It begins by saying not to use the features in a way that violates applicable laws or harms others, and then says “for example, you must not”:

1. Create explicit sexual material or explicit nudity.
2. Create hateful or highly offensive content that attacks or dehumanises a group based on race, ethnicity, national origin, religion, serious disease or disability, gender, age or sexual orientation.
3. Glorify graphic violence or gore.
4. Promote self-harm.
5. Depict nude minors or minors in a sexual manner.
6. Promote terrorism or violent extremism.
7. Spread misleading, fraudulent or deceptive content that could lead to real-world harm.
8. Violate others’ privacy rights.
9. Engage in regulated activities without complying with applicable requirements.

The first item is the one that bears on adult games. The list is introduced as examples, and the child safety item (the fifth) is separate from the adult content item.

What the wording does and does not say

Three observations follow from the text.

• The standard is explicitness. The words are “explicit” for both sexual material and nudity. The guidelines do not mention suggestive, revealing or provocative content, so those are not addressed one way or the other.
• The list is “for example”. The word “example” means the list is not exhaustive. Adobe’s reserved right to disable accounts for any violation applies to the whole document, not just the nine items.
• No exception is stated. There is no allowance for adult accounts, artistic context or age verification.

Deceptive behaviour

A second list in the same section covers deceptive or harmful behaviour: using fake or inaccurate profile information, impersonating people or entities, using unauthorised automated or scripting processes such as bulk uploading through a script, and schemes to boost engagement metrics. It is a reminder that the guidelines cover account conduct as well as content.

Review and reporting

The guidelines say prompts and inputs to generative AI features in Adobe products, and the results those features generate, may be reviewed through both automated methods, such as machine learning, and manual methods, for abuse prevention and content filtering. They give an address, [email protected], for reporting a suspected violation, and say Adobe may report any material exploiting minors to the National Center for Missing & Exploited Children.

The statement is about review, not about what is reviewed in every case. The guidelines do not say how often prompts are read by a person, or what triggers a manual look.

Third-party rights

The second section covers other people’s rights. Using the features to create, upload or share content that violates third-party copyright, trademark, privacy, publicity or other rights is prohibited. The guidelines give examples of what that may include:

• entering text prompts designed to generate copyrighted, trademarked or otherwise infringing content;
• uploading an input or reference image that includes a third party’s copyrighted content;
• generating text that plagiarises third-party content;
• using a third party’s personal information in violation of their privacy or data protection rights.

For a game artist, the reference image rule is the practical one. Feeding a screenshot of another studio’s character into a generative tool as a reference is on the list of examples. The section also sends rights holders to Adobe’s Intellectual Property Removal Policy and says Adobe cannot moderate contract or employment disputes between users.

Customised features

A section on customised generative AI features applies to people who build and share their own versions of a feature, such as an assistant. The guidelines make the builder responsible for how it is configured, deployed and used, and say the customised feature must not be configured to do certain things, including simulating a human-like emotional relationship with a user. That item matters to a developer who might consider building a companion-style character on top of a generative feature: the guidelines list it as something a customised feature must not do.

Content Credentials

The last numbered section covers Content Credentials. Adobe says it may attach or publish them for content created or modified with generative AI features, to let people know the content was generated with or modified by AI. Users must not remove, alter or disable Content Credentials.

For game developers, this links to store rules on AI disclosure. A store that asks you to label AI-assisted assets, covered in does itch.io allow AI-generated games and AI-generated content rules for adult games, will find the same fact in the file’s metadata if the credentials are intact.

Other sections

Section 4, Use Your Judgment, says outputs may be inaccurate or misleading and that quality may be reduced for languages Adobe does not explicitly support, and it tells users not to enter sensitive personal information unless the product is designed for it. Section 5 says outputs are not professional advice.

What this means for adult game developers

1. Do not generate explicit sexual art or explicit nudity with Adobe’s generative features. It heads the list of prohibited uses.
2. Assume prompts and results can be read. Review by automated and manual methods is stated.
3. Keep character ages unambiguous. The guidelines bar nude minors and minors in a sexual manner.
4. Do not use other studios’ art as reference input. Copyrighted reference images are a listed example.
5. Leave Content Credentials alone. Removing them is prohibited, and they may be the evidence a store wants for AI disclosure.
6. Do not build a companion-style persona on a customised feature. Simulating a human-like emotional relationship is a listed restriction.
7. Keep non-generative editing separate. The guidelines govern the generative features. Whether ordinary editing tools fall under other terms was not read.
8. Check where the art will be sold. Store rules differ. See where to sell adult games, and the Journal’s guide to Canva’s rules for a comparable design tool.

What is still unknown

• How the filters treat suggestive or revealing art. The guidelines do not mention it.
• Which products the guidelines cover. The text read does not list them.
• How often manual review happens. The guidelines state that it may.
• What account action looks like in practice. The guidelines state a right, not a ladder.
• What Adobe’s terms of use add. They were not read.

How this page was put together

The Generative AI User Guidelines were read on Adobe’s own site, in a browser, on 4 November 2026. The page shows a last updated date of 15 May 2026. Short phrases are quoted and the rest is paraphrased.

The author has not tested what Firefly or Photoshop’s generative features will and will not produce, and nothing here is a first-hand account of their filters. Adobe community forum discussions of filter behaviour appeared in a search but were not read, and Adobe’s general terms of use and the product-specific pages were not read.

This page was written on 4 November 2026 and will be updated when Adobe changes its guidelines. To report an error, see the corrections standard.