We build tools that create. We hold the line on how they're used.
Tagshop AI generates video ads at scale. That power comes with real responsibilities — to the people whose faces appear on camera, to the brands who publish, and to the audiences who watch.
Four principles that govern everything we build
These aren't aspirations written for a policy page. They're the constraints that shape how we design features, approve avatars, review content, and respond to reports.
Consent first, always
No face, voice, or likeness enters the Tagshop AI platform without the documented, explicit consent of the real person it belongs to. Consent governs avatars, AI Twins, and every creative input. It can be withdrawn, and we remove the relevant content within 48 hours when it is.
Transparency, not concealment
Every video Tagshop AI generates carries provenance metadata — a permanent, machine-readable record that it was created by artificial intelligence. We support C2PA content credentials and embed invisible watermarks on all paid-plan outputs. We do not help users hide that their content is AI-generated.
Safety built in, not bolted on
Content moderation classifiers run on every prompt and every output before it reaches a user. Safety is not a layer applied after generation — it is part of the generation pipeline. We maintain a dedicated Trust and Safety team that updates these systems as threats evolve.
Humans stay accountable
Automated systems flag; people decide. Every account action, every report response, every avatar removal is reviewed by a human team member. We respond to all reports within 48 hours. Where the law requires it, we report misuse to relevant authorities — without requiring a court order in cases involving child safety.
Every face here belongs to a real person who said yes
Tagshop AI's 300+ avatar library was built through a deliberate consent process. The same applies to every AI Twin created on our platform. This section explains exactly what that means.
Library avatars are created from real actors and models who signed an explicit agreement allowing their likeness to be used in AI-generated commercial video content. That agreement specifies the permitted uses, the compensation terms, and their right to withdraw at any time. We do not acquire avatars from stock imagery, scraped datasets, or any source where the subject did not specifically consent to AI video use.
The AI Twin feature allows users to create a digital clone of their own face and voice — and only their own. We require identity verification before AI Twin activation. Generating a twin of another person — public figure, colleague, creator, or anyone else — without their documented consent is a Terms of Service violation that results in immediate account suspension.
Every avatar in our library has a signed consent record. We retain these records and can produce them in response to a valid legal request or verified withdrawal of consent.
An avatar holder can withdraw consent at any time by contacting [email protected]. The avatar is removed from the platform within 48 hours and is no longer available for new generations.
Before generating an AI Twin, users complete an identity verification step that confirms the face they are cloning matches the identity on their account. This is not optional.
Uploading images or audio of another person to generate their likeness is prohibited and results in immediate account suspension. This applies to public figures, celebrities, and private individuals alike.
What this platform is for — and what it is not
Violations result in immediate content removal, account suspension, and — where legally required — reporting to law enforcement or the National Center for Missing and Exploited Children.
What we enforce at the infrastructure level
These safeguards run on every generation on our platform. They are not visible to users — and that is intentional.
Content moderation on every prompt and output
Automated classifiers screen both the input prompt and the generated output before delivery. They are trained on prohibited content categories including NCII, hate speech, child safety violations, impersonation signals, and violent extremism. Flagged content is blocked and queued for human review.
C2PA content credentials — provenance by default
All Tagshop AI outputs carry Content Authenticity Initiative (C2PA) metadata identifying them as AI-generated. This provenance record is embedded in the file itself and persists through most common editing, compression, and re-upload workflows. We are an active supporter of the C2PA standard.
Invisible watermarking on all paid outputs
A non-visible watermark is embedded in every video generated on a paid Tagshop AI plan. Unlike metadata, this watermark is designed to survive transcoding, cropping, and re-encoding. It allows us — and platform partners — to identify Tagshop AI content even when metadata has been stripped.
Identity verification before AI Twin activation
Users cannot activate the AI Twin face-cloning feature until they complete an identity verification step that confirms the face they are uploading matches their account identity. This check is mandatory and cannot be bypassed.
Human review of all flagged content within 24 hours
Content that triggers automated flags is placed in a review queue. A human Trust and Safety team member reviews every item in this queue within 24 hours. Account actions — warnings, suspensions, terminations — are always taken by people, not only by algorithms.
Rate limiting and anomalous usage detection
API and generation endpoints are rate-limited per account. Anomalous usage patterns — such as generating the same avatar at high volume, repeated re-upload of removed content, or generation velocity inconsistent with stated use cases — trigger automatic review queues and, where warranted, account holds.
Your content and your data are yours
We do not sell, rent, or share your personal data with third parties for advertising or profiling. Your videos, prompts, and account information are stored securely and deleted on request.
Encryption in transit and at rest
All data transmitted to and from Tagshop AI is encrypted using TLS 1.3. Data at rest is encrypted with AES-256.
GDPR and CCPA compliant
We comply with GDPR for users in the European Union and CCPA for users in California. Your privacy rights are honoured regardless of your location.
Data deletion on request
You can request deletion of your account and all associated data at any time. Deletion is processed within 30 days. Contact [email protected].
6-month video retention
Generated videos are retained for 6 months on standard plans and are automatically deleted at expiry unless downloaded or saved by the user.
No training on your content
Your videos, prompts, and creative assets are not used to train Tagshop AI's models without your explicit opt-in. You can review your data settings at any time.
Subprocessor transparency
A complete list of subprocessors — the third-party services that process data on our behalf — is maintained and updated at tagshop.ai/privacy.
Running AI-generated ads responsibly
Tagshop AI is purpose-built for advertising. That puts us — and our users — under a higher standard than general-purpose AI tools, and we take that seriously.
The FTC's guidelines on AI-generated advertising content continue to evolve. Our current guidance: disclose that your ad is AI-generated wherever the platform requires it, and when in doubt, err toward disclosure. Tagshop AI does not provide legal advice, and you are responsible for ensuring your ads comply with applicable laws and platform rules in the markets where you run them.
FTC disclosure
AI-generated video ads must comply with FTC guidelines on disclosure. Tagshop AI templates include guidance on disclosure language. You are responsible for final compliance with the rules of every platform and jurisdiction in which you run ads.
No brand or trademark impersonation
Creating ads that impersonate another brand, misuse trademarks, or create a false impression of endorsement is prohibited and may constitute a legal violation. This applies regardless of how the content is labelled.
Financial and health claims
Exaggerated earnings claims, unsubstantiated health claims, and misleading before/after comparisons are prohibited in Tagshop AI-generated content — regardless of whether they would otherwise comply with platform rules. This applies to supplements, financial products, weight loss, and similar high-risk categories.
Commercial licence scope
The commercial licence included on paid plans covers running AI-generated content as paid advertising on Meta, TikTok, Google, and other platforms. It does not extend to sublicensing individual generated videos to third parties, or using Tagshop AI avatars in contexts outside the platform without a written enterprise agreement.
Something wrong? Tell us.
You do not need a Tagshop AI account to submit a report. We review every submission and respond within 48 hours. In urgent cases involving child safety, we escalate to relevant authorities without waiting for that window.
Submit a report
Describe what you encountered and we'll investigate. The more specific you are, the faster we can act.