MiraFrame AI is an image and video generation product for lawful creative work. This policy explains the safety rules that apply to prompts, uploads, reference images, generated images, generated videos, edits, templates, API usage, and account activity.
By using MiraFrame, you agree to use the service only for legitimate, ethical, and rights-respecting purposes.
AI product disclosure
MiraFrame uses third-party AI model providers and related infrastructure to process generation requests. AI outputs may be unexpected, inaccurate, incomplete, or unsuitable for a particular commercial use. You are responsible for reviewing outputs before publishing, selling, advertising, or otherwise distributing them.
Generated content may include a MiraFrame AI watermark or other product labeling. Do not remove, hide, or misrepresent AI-origin labels where doing so would be deceptive, unlawful, or against provider rules.
User responsibility and rights
You are responsible for the content you submit and the outputs you choose to use. You must have all necessary rights, permissions, licenses, and consents for prompts, uploads, reference images, names, brands, faces, likenesses, copyrighted works, trademarks, and any other material you provide to MiraFrame.
Do not use MiraFrame to create or distribute content that infringes copyright, trademark, publicity rights, privacy rights, contractual rights, or other intellectual property or personal rights.
Prohibited adult and NSFW content
MiraFrame does not allow adult, pornographic, sexually explicit, nude, fetish, sexual service, or other NSFW generation workflows. This includes attempts to create sexualized images or videos of real people, fictional minors, young-looking people, or any person who has not clearly consented.
We may block, remove, refuse to process, or restrict accounts involved in sexual exploitation, non-consensual intimate imagery, sexualized minors, or attempts to bypass these restrictions.
Prohibited violence, gore, and self-harm content
MiraFrame does not allow graphic violence, gore, torture, mutilation, violent-extremist propaganda, glorification of real-world violence, or instructions that enable violent wrongdoing. Content that encourages, instructs, or glorifies self-harm or suicide is also prohibited.
Fictional, educational, historical, or news-related depictions may still be restricted when they are excessively graphic, exploitative, or likely to cause harm. We may require a safer framing, refuse the request, or remove the output.
Prohibited hate, harassment, and discrimination
Do not use MiraFrame to dehumanize, threaten, harass, discriminate against, or promote violence against a person or group based on protected characteristics. This includes hateful symbols or propaganda used to intimidate, target, recruit, or celebrate abuse.
Child safety and CSAM
Any child sexual abuse material (CSAM), sexualization of minors, grooming, exploitation, or unsafe sexual depiction of minors is strictly prohibited. We prioritize credible child-safety reports and may preserve relevant information and report it to appropriate providers or authorities where required by applicable law.
Face, likeness, deepfake, and impersonation restrictions
MiraFrame must not be used for deceptive or non-consensual identity manipulation. Prohibited uses include:
- face-swap, deepfake, or face manipulation workflows without clear rights and consent;
- impersonating a real person, public figure, brand, employee, government representative, or service provider;
- creating misleading endorsements, testimonials, profile images, documents, or evidence;
- generating non-consensual intimate, defamatory, harassing, or exploitative likeness content.
For commercial, public, or advertising use involving a person's likeness, you are responsible for obtaining appropriate consent and reviewing applicable laws.
Copyright, trademark, and IP restrictions
Do not upload, reference, generate, or distribute material that you do not have the right to use. This includes copyrighted characters, artwork, logos, brand assets, celebrity likenesses, protected designs, unreleased media, paid stock images, and other third-party assets unless you have the necessary authorization.
Do not use MiraFrame to mislead consumers into believing that your output is affiliated with, endorsed by, or produced by another company, creator, artist, platform, or rights holder.
High-risk and harmful uses
MiraFrame may not be used for illegal, deceptive, abusive, or high-risk activity, including:
- fraud, scams, phishing, spam, fake ads, fake reviews, or deceptive marketing;
- harassment, threats, hate, discrimination, bullying, or targeted abuse;
- malware, credential theft, security evasion, or instructions for wrongdoing;
- political manipulation, voter deception, false evidence, or synthetic news designed to mislead;
- medical, legal, financial, identity, government, safety-critical, or other high-stakes decisions;
- content that promotes self-harm, violence, exploitation, weapons misuse, or unlawful activity.
Safety controls and moderation
MiraFrame uses automated moderation checks for AI image and video prompts before generation, including Creem Moderation where enabled for payment and compliance review. We may also use provider safety systems, rate limits, manual review, account restrictions, content removal, and other safeguards to protect users and comply with legal, platform, and payment requirements. Some requests may be refused even if similar content appears elsewhere online.
Where required for platform, payment, or compliance review, MiraFrame may add or update third-party moderation tools and additional verification steps.
Our moderation process combines automated screening with risk-based human review. Automated systems may inspect prompts, uploaded inputs, and generation requests before or during processing. Flagged requests may be rejected, rate-limited, held for review, removed, or escalated. We investigate credible user reports and use the information reasonably necessary to assess the reported content or account activity.
Current security practices
Encrypted transport
Public and authenticated MiraFrame pages are served over HTTPS. This protects data in transit between the browser and the service, but it does not make unsafe passwords or unauthorized uploads secure.
Private account routes
Account settings, billing, credits, generation activity, administration, and API routes are excluded from public search indexing. Authentication and permission checks are applied to protected workflows where required.
Restricted browser capabilities
MiraFrame sends security-related response headers that restrict framing, MIME-type guessing, referrer information, and unnecessary access to camera, microphone, and location features.
Limited data sharing
Generation requests may be transmitted to the model provider selected by the workflow. Users should avoid uploading confidential information or third-party material unless they are authorized to process it under the applicable provider terms.
Protect your account
- Use a unique password that is not reused on another service.
- Protect access to the email account used for sign-in.
- Never share verification codes, session details, or private API keys.
- Review unexpected account, billing, or generation activity promptly.
- Sign out on shared or public devices.
Reporting policy, rights, or security concerns
Send suspected policy violations, copyright concerns, likeness concerns, or security issues to [email protected].
For content or abuse reports, use the subject Content Safety Report and include the relevant URL, task or case number, screenshots, a description of the concern, and any rights or authority evidence. We review reports as reasonably practicable and prioritize credible reports involving child safety, imminent threats, unlawful content, or serious impersonation.
For security reports, use the subject Security report and include:
- the affected URL or workflow;
- clear reproduction steps;
- the expected and observed behavior;
- the potential impact;
- screenshots or logs with secrets and personal data removed.
For copyright, trademark, or likeness concerns, include the affected URL or asset, a description of the rights concern, and contact information so we can follow up.
Do not exploit an issue beyond what is necessary to demonstrate it. Do not access other users' data, disrupt the service, run destructive tests, or publicly disclose an unresolved issue.
Enforcement
MiraFrame may refuse prompts, block uploads, remove outputs, cancel tasks, disable features, limit model access, limit or freeze credits or payments where appropriate, suspend accounts, terminate accounts, or preserve information where needed for abuse prevention, legal compliance, provider obligations, payment compliance, or platform safety.
Enforcement is based on the severity of the conduct, user intent, repeat behavior, evidence available, and applicable law. We may issue a warning for lower-risk violations, while serious or repeated violations can result in immediate restrictions or termination. When legally required or necessary to protect users and partners, we may report matters to model providers, payment partners, or relevant authorities. Users may request reconsideration by emailing [email protected] with the task ID and relevant context; restrictions may remain in place while we review the request.
Scope and limitations
This page describes current practices and is not a certification, warranty, or guarantee that every security, policy, or content risk can be prevented. Safety controls evolve as the product, providers, and infrastructure change. The Privacy Policy and Terms of Service remain the governing policy documents.