Responsible AI
CINEM8E combines human creative direction with AI-assisted story analysis, image generation, video generation and continuity tools. Responsible use requires lawful inputs, informed consent, human review and transparent treatment of synthetic media.
1. Human creative control
CINEM8E is designed to assist—not replace—writers, directors, producers, editors, rights holders and safety reviewers. Users remain accountable for instructions, selections, approvals and final publication.
2. Transparency
Where context requires it, disclose that media is AI-generated or materially AI-modified. Do not use CINEM8E outputs to fabricate evidence, impersonate a person or create false endorsements.
3. Consent and identity
Obtain valid consent and authority before using a person’s face, voice, performance, private information or distinctive identity. Reference images must not be used to create non-consensual intimate, defamatory or exploitative media.
4. Children and vulnerable people
Use heightened safeguards for minors and vulnerable individuals. Do not create sexualised, exploitative, coercive, humiliating or dangerous depictions, or upload their personal data without lawful authority and appropriate consent.
5. Rights-respecting creation
Upload only material you are authorised to use. Respect copyright, trademarks, publicity rights, privacy, cultural heritage and contractual restrictions. Generated output must undergo clearance before commercial release.
6. Safety restrictions
Do not use CINEM8E for unlawful instructions, targeted harassment, hate or dehumanisation, sexual exploitation, violent wrongdoing, fraud, political manipulation, malicious deepfakes, evasion of safeguards or other material harm.
7. Fairness and representation
AI systems can reproduce stereotypes and uneven performance. Review characters, language, casting, culture and visual treatment for bias, disrespect or exclusion. Correct problematic output rather than relying on automated generation.
8. Accuracy and provenance
Do not rely on generated content as factual, legal, medical, financial or historical evidence. Maintain project records and provenance appropriate to the production, especially when output represents real events or people.
9. Privacy and minimisation
Use the least personal or confidential information necessary. Avoid secrets, biometric material and sensitive records unless processing is authorised, contractually protected and essential to the production.
10. Security and access
Use individual credentials, restrict workspace access and report suspected compromise. Do not attempt to bypass controls, extract proprietary prompts, probe other projects or interfere with platform infrastructure.
11. Model limitations
Generation may produce continuity errors, unsafe details, artifacts, inaccurate dialogue or unintended resemblance. Outputs must be inspected frame-by-frame and corrected before distribution.
12. Enforcement
Music Gen Pvt. Ltd. may investigate suspected misuse, restrict generation, preserve relevant security records, suspend accounts or cooperate with lawful requests where necessary to protect users, rights holders and the public.
13. Report a concern
Report harmful output, non-consensual identity use, suspected infringement, security issues or Responsible AI concerns to support@themusicgen.com. Include the project identifier and relevant context without resending unnecessary sensitive material.