AI-Generated Content Policy
Last updated: June 2026
1. Synthetic Origin — No Real Person Depicted
Every face and image produced by Tasvr, except where you supply your own or a consented third party's reference photo, is algorithmically synthesized from statistical patterns learned during model training. Output is not a photograph, scan, digital replica, or composite of any specific, identifiable living or deceased individual, and no output is created with the intent to depict, represent, or endorse any real person.
2. Coincidental Resemblance
Because our models are trained on large, diverse datasets and tuned to represent a wide range of Indian ethnic and regional facial features, an output may, by statistical coincidence, resemble a real person. Any such resemblance is unintended and not a representation, depiction, or endorsement of that person. We do not grant, and cannot grant, any right of publicity or likeness clearance in outputs, because they aren't derived from any real person's protected likeness.
3. Reporting a Resemblance Concern
If you believe a specific output identifiably resembles a real, living individual, email tasvrindia@gmail.com. We review reports in good faith and, where warranted, remove or regenerate the asset and notify affected licensees.
4. AI-Content Labeling
Consistent with India's IT Rules on synthetically generated information and the EU AI Act's transparency requirements, photorealistic output is machine-marked as AI-generated at the point of generation, via a visible label and/or embedded provenance metadata. You must not strip, disable, or misrepresent this marking, and you're responsible for any additional disclosure required by law where you use the output (for example, advertising-disclosure rules for synthetic performers).
5. If You Modify Output to Target a Real Person
If you edit, retouch, or combine output — or use your own reference-photo upload — to more closely target or identify a specific real person, that use falls outside this policy's disclaimer and outside our indemnification. You're solely responsible for any consents or clearances that use requires and for compliance with right-of-publicity and deepfake laws that may apply.
6. Prohibited Uses
You may not use Tasvr or its output to impersonate or falsely suggest endorsement by a real person, create content that a reasonable viewer would believe depicts a real identifiable person in a defamatory, sexual, medical, criminal, or otherwise sensitive context, or circumvent any digital-replica, right-of-publicity, or synthetic-media-labeling law. See our Acceptable Use Policy for the full list.
7. No Warranty of Non-Resemblance
We warrant that output is algorithmically generated and, to the best of our knowledge, not intentionally modeled on any specific identifiable real individual. We do not warrant that no output will ever coincidentally resemble a real person. If you intend to use output in a sensitive context — health, legal, financial, political, or advertising — we recommend an independent visual-similarity check before wide publication.
8. Indemnification
See our Terms of Service for the full indemnification structure. In summary: we'll support you against right-of-publicity claims arising solely from unmodified, license-compliant use of output that coincidentally resembles a real person, provided you notify us promptly and stop using the specific asset on request. Modifications or misuse as described in sections 5–6 shift responsibility to you.