Acceptable Use Policy

This Acceptable Use Policy ("AUP") defines what you can and cannot do with Depicta (depicta.ai), operated by 30ohm L.P.. It applies to all users and all methods of access — API, CLI, MCP server, skill files, and dashboard.

This policy is part of our Terms of Service. Violations may result in content removal, account suspension, or permanent termination.

1. Our Commitment

We believe in enabling creativity while preventing harm. Our content moderation is designed to stop genuinely harmful content without unnecessarily restricting legitimate creative use. If you believe we've made an error, we have an appeals process (see Section 6).

2. Prohibited Content

You must not use Depicta to generate, upload, edit, or process any of the following:

2.1 Absolute Prohibitions (Zero Tolerance)

These result in immediate account termination and referral to law enforcement. No warnings, no appeals on the substance.

  • Child sexual abuse material (CSAM) or any sexualization of minors — including fictional, cartoon, or AI-generated depictions. This is a criminal offense in all EU Member States.
  • Non-consensual intimate imagery — sexually explicit content depicting real, identifiable individuals without their documented consent, including AI-generated depictions ("deepfakes")
  • Terrorist content — material that incites, promotes, or glorifies terrorism, or provides instructions for carrying out terrorist acts

2.2 Prohibited Content Categories

These result in content rejection and graduated enforcement (see Section 5).

  • Incitement to violence — content that encourages, threatens, or glorifies violence against specific individuals or groups
  • Hate speech — content that attacks, demeans, or dehumanizes individuals or groups based on race, ethnicity, national origin, religion, gender, sexual orientation, disability, or other protected characteristics. This includes Holocaust denial, genocide denial, and gross trivialization of genocide, crimes against humanity, or war crimes — which are criminal offenses under EU Framework Decision 2008/913/JHA and Greek Law 927/1979
  • Self-harm and suicide — content that promotes, instructs, or glamorizes self-harm or suicide as a positive act or solution. This does not include mental health awareness imagery, prevention campaigns, or artistic exploration of mental health themes
  • Harassment and threats — content targeting specific individuals with intimidation, stalking imagery, or threatening material
  • Unauthorized depictions of real people — photorealistic generation of identifiable living persons without their consent. This restriction exists because generating realistic imagery of a person may constitute an offense under Greek personality rights law (Civil Code Art. 57) and EU data protection law, even without publication. It does not apply to caricature, cartoon, or clearly artistic depictions of public figures (which are protected expression)
  • Identity fraud and impersonation — content created to impersonate real individuals for deceptive purposes, including fake identity documents, forged communications, or misleading deepfakes
  • Intellectual property infringement — generating, uploading, or editing content that infringes the intellectual property rights of others, including prompts designed to induce the service to reproduce protected works or identifiable protected material. This does not apply to fair use, parody, quotation for criticism or review, education, research, news reporting, or other uses permitted under applicable law — but you bear sole responsibility for ensuring your use qualifies (see Terms of Service, Section 7.3)
  • Illegal content — content whose creation, possession, or distribution is illegal under the laws of Greece, the European Union, or the user's jurisdiction
  • Dangerous instructions — content that provides actionable instructions for creating weapons, drugs, or other items intended to cause physical harm

2.3 Special Category Personal Data

Depicta does not solicit, require, or rely on special category personal data (as defined in GDPR Article 9) for any processing purpose. You must not submit prompts or uploaded content that include such data about yourself or any identifiable third party. Special category data includes personal data revealing:

  • Racial or ethnic origin
  • Political opinions
  • Religious or philosophical beliefs
  • Trade union membership
  • Genetic data, or biometric data processed for the purpose of uniquely identifying a natural person
  • Data concerning health
  • Data concerning a natural person's sex life or sexual orientation

If you need to generate imagery for a medical, religious, political, or similarly sensitive context, describe the scene in neutral, non-identifying terms without referencing specific individuals or their attributes. You bear full responsibility for ensuring that your prompts and uploads comply with this requirement. Depicta's content classifier is not designed to detect special category data as such — this restriction is enforced contractually, not technically.

3. Prohibited Conduct

Beyond content restrictions, the following uses of Depicta are prohibited:

  • Circumventing safety measures — attempting to bypass, disable, or trick the content classifier through prompt engineering, obfuscation, adversarial inputs, or any other technique
  • Automated abuse — using bots, scripts, or automated systems to submit large volumes of requests designed to exploit the service, test content moderation boundaries, or degrade service quality
  • Credential sharing — sharing your account credentials or API keys with others, or using another person's credentials
  • Multiple accounts — creating multiple accounts to circumvent rate limits, bans, usage restrictions, or trial limits
  • Unauthorized security testing — scanning for vulnerabilities, probing API endpoints for weaknesses, fuzzing inputs, attempting to exploit bugs or misconfigurations, or conducting any form of penetration testing without our prior written authorization. If you discover a vulnerability, report it to security@depicta.ai — do not attempt to exploit or verify it.
  • Unauthorized access — accessing or attempting to access accounts, systems, data, or infrastructure that you are not authorized to use, including through credential stuffing, brute-force attacks, session hijacking, or exploitation of software vulnerabilities
  • Reverse engineering — decompiling, disassembling, or reverse-engineering Depicta's software, except as permitted by EU Directive 2009/24/EC
  • Competitive misuse — using Depicta to build a competing service, or systematically extracting data about our models, routing, or prompt engineering
  • Fraudulent activity — using the service in connection with fraud, phishing, scams, or other deceptive practices
  • Removing provenance markers — stripping, altering, or disabling C2PA Content Credentials or digital watermarks embedded in generated images

4. Content Moderation

4.1 How It Works

All prompts and uploaded images are screened by an automated content classifier before processing. The classifier operates independently of the generation pipeline — it runs in an isolated environment, making binary allow/deny decisions.

  • Allowed requests proceed to image generation
  • Denied requests are rejected before any image is generated. You are not charged for denied requests.

4.2 What We Log

When the classifier denies a request, we log:

  • The classifier's decision category (e.g., "violence", "CSAM")
  • A timestamp
  • The associated user ID

Denied request logs are retained for 12 months and used for safety auditing and pattern detection.

4.3 Limitations

No automated system is perfect. The classifier may occasionally:

  • False positive — reject a legitimate prompt. If this happens, you can appeal (see Section 6).
  • False negative — fail to catch prohibited content. If you encounter content that should have been caught, please report it (see Section 7).

5. Violations and Consequences

5.1 Serious Violations

For absolute prohibitions (Section 2.1: CSAM, non-consensual intimate imagery, terrorist content) and serious security violations (unauthorized access, exploitation of vulnerabilities, or other conduct that compromises the security or integrity of Depicta):

  1. Immediate account termination — no prior warning
  2. Forfeiture of remaining credits
  3. Referral to law enforcement in the relevant jurisdiction
  4. Preservation of evidence for legal proceedings
  5. Permanent ban — you may not create a new account

5.2 Other Violations

For prohibited content (Section 2.2) and prohibited conduct (Section 3), we follow a graduated response:

  1. First violation: Warning via email with a clear description of the violation and the specific policy section breached
  2. Second violation: Temporary account suspension (duration depends on severity; typically 7–30 days)
  3. Third violation or severe conduct: Permanent account termination and forfeiture of remaining credits

We may skip steps in the graduated process if the violation is severe, involves clear malicious intent, or poses an immediate risk to safety.

5.3 During Suspension

While your account is suspended:

  • You cannot make API calls, generate images, or access the service
  • Your account data and credit balance are preserved
  • You retain access to your dashboard to view account status and contact support

6. Appeals

6.1 When You Can Appeal

You may appeal:

  • A content moderation decision (classifier denial) that you believe is incorrect
  • An account suspension or termination that you believe is unjustified
  • A warning that you believe was issued in error

You may not appeal the substance of an immediate termination under Section 5.1 (absolute prohibitions). You may appeal only if you believe the factual basis was incorrect (e.g., the flagged content was misidentified).

6.2 How to Appeal

Send an email to hello@depicta.ai with:

  • Subject line: "AUP Appeal"
  • Your account email address
  • A description of the decision you are appealing
  • Why you believe the decision was incorrect

6.3 Review Process

  • Appeals are reviewed by a human — not the automated classifier
  • We aim to respond within 5 business days
  • If we uphold the decision, we will explain why
  • If we reverse the decision, your access will be restored and any affected credits refunded

7. Reporting Content

If you encounter content generated through Depicta that you believe violates this policy, or if you discover that our classifier has failed to catch prohibited content, please report it:

  • Email: hello@depicta.ai
  • Subject line: "Content Report"
  • Include as much detail as possible: the content in question, how you encountered it, and why you believe it violates this policy

We take all reports seriously and will investigate promptly.

8. Changes to This Policy

We may update this Acceptable Use Policy from time to time.

Material changes — changes that add new categories of prohibited content, alter enforcement consequences, or reduce your rights — will be communicated with at least 30 days' notice via email and on the Depicta website.

Non-material changes — such as clarifications that do not alter the scope of existing prohibitions, corrections of typographical errors, or formatting adjustments — take effect when posted on the website.

The updated policy will display the new "Last updated" date.

9. Contact

If you have questions about this policy: