AI Usage & Personal Data Policy
How ZorgSocial uses artificial intelligence in our iOS and Android apps, which AI models process your content, and how your camera, photo gallery, microphone and other permissions control exactly what is shared.
Last Updated: September 13, 2026
Our commitment to transparent, permission-based AI
ZorgSocial is an AI-powered marketing platform. Our iOS and Android applications use artificial intelligence to generate and edit images, create and edit video, produce voiceovers, music and dubbing, transcribe and translate audio, and to power the Easy Zorg agent that plans and executes multi-step creative work for you. This policy explains, in plain language, how that works, what data leaves your device, which AI models receive it, and how you stay in control.
This policy supplements β and should be read together with β our Privacy Policy, our AI Terms of Use, our Acceptable Use Rules and the AI Sub-Processors & Providers list. Where this policy is more specific about AI processing in the mobile apps, it prevails for that subject matter.
The short version
Nothing is sent to any AI model unless you have granted the relevant device permission (camera, photo gallery, microphone or files) and given your explicit in-app consent for third-party AI processing. You can withdraw consent and revoke permissions at any time, and you can read the privacy policy of every model provider we use from the inventory below.
1. Scope of this Policy
This policy applies to the ZorgSocial mobile applications for iOS and Android, the ZorgSocial web application, and any AI-powered feature made available through them (together, the "AI Features"). It covers personal data and content that you provide to us or that we process on your behalf when you use those AI Features.
The AI Features include, without limitation:
- Image generation and image editing, including reference-based editing, portrait retouching, background removal, upscaling and vector (SVG) generation
- Video generation, image-to-video animation, video editing, video extension, upscaling and lip-sync
- Text-to-speech, voice cloning and voice changing, multi-speaker dialogue, music generation, sound design and AI dubbing / speech translation
- Transcription, translation, captioning and audio understanding
- The Easy Zorg agent, which reasons about your request, plans multi-step workflows and selects the appropriate models and tools
- Content generation, copywriting, hashtag and caption generation, strategy planning, analytics and social listening
2. Device Permissions: Your First Line of Control
Our mobile apps request operating-system permissions only when a feature needs them, and only for the purpose described. The table below explains what each permission is used for, what it does not do, and how you can change it.
Camera
Used when you choose to capture a photo or record a video directly in the app β for example to create a product shot, a talking-head avatar, a reference image for generation, or a video clip to edit. The camera is only active while you are actively capturing; there is no background or passive camera access.
Your control: You can deny or revoke camera access in iOS Settings β Privacy & Security β Camera β ZorgSocial, or Android Settings β Apps β ZorgSocial β Permissions β Camera. If you deny it, you can still upload existing media from your gallery instead.
Photo Gallery / Media Library
Used when you select existing photos or videos to upload as inputs, references or assets for an AI Feature. On iOS you may grant access to selected items only, or to your full library; on Android you may grant access to selected photos and videos only. We only read the specific items you select.
Your control: You can change this at any time in iOS Settings β Privacy & Security β Photos β ZorgSocial (choose "Selected Photos" or "None"), or Android Settings β Apps β ZorgSocial β Permissions β Photos and videos. Revoking access does not delete content you previously uploaded.
Microphone
Used when you record a voiceover, dictate a prompt, capture a voice sample for voice cloning or voice changing, or record audio to be transcribed, translated or dubbed. Recording only occurs while you are actively recording in the app.
Your control: You can deny or revoke microphone access in iOS Settings β Privacy & Security β Microphone β ZorgSocial, or Android Settings β Apps β ZorgSocial β Permissions β Microphone. You can still type prompts and upload existing audio files instead.
Files & Documents
Used when you attach a PDF, document, spreadsheet or other file as an input or reference for an AI Feature, or when you import media from cloud storage. We access only the files you explicitly pick.
Your control: File access is granted per selection through the system file picker. You can stop granting access simply by not selecting files; you can also remove previously imported files from your ZorgSocial asset library at any time.
Notifications
Used to tell you when a long-running generation job has finished, when a scheduled post is about to publish, or when the Easy Zorg agent needs your approval to continue a workflow. Notifications do not carry your media content.
Your control: You can disable notifications in iOS Settings β Notifications β ZorgSocial, or Android Settings β Apps β ZorgSocial β Notifications, without affecting any AI Feature.
Location
ZorgSocial does not require precise location for AI Features. If you optionally add a location to a post or campaign, you type or select it yourself. We do not collect background location.
Your control: You can deny or revoke location access in your device settings. AI Features continue to work normally without it.
Denying a permission never blocks the rest of the app. It only disables the specific capability that depends on it. Every AI Feature that needs media also accepts an upload from your gallery or files, so you always have an alternative path.
On both iOS and Android, the operating system itself enforces these permissions. Even if ZorgSocial were to request access, the OS will not deliver camera, photo, microphone or file data unless you have granted it. You can review and revoke permissions at any time without deleting your account.
3. How Your Content Reaches an AI Model
When you use an AI Feature, your content follows a controlled path. At each step, the app tells you what will happen before it happens:
- You select or capture media, or type a prompt, inside the app.
- The app shows a just-in-time notice identifying the AI Feature, the type of content to be processed, and β where the model is known β the AI model and provider that will process it.
- You approve the action. No content is transmitted before this approval.
- Your content is uploaded over an encrypted connection (TLS) to ZorgSocial's own infrastructure, where it is stored so you can access your job history and reuse sessions.
- ZorgSocial submits the content to the selected third-party AI model provider using ZorgSocial's own service credentials. The provider does not receive your ZorgSocial account identity, name, email address or payment details.
- The provider returns the generated or edited output, which is stored on ZorgSocial's infrastructure and shown to you in the app.
- You can delete the job, the session or your account at any time, which removes the associated content from our systems subject to the retention rules in Section 9.
What may be sent to an AI model
- Images and photographs you upload or capture, including photographs that may contain identifiable faces where you have chosen to use a portrait, avatar or face-editing feature
- Videos and video clips you upload, capture or generate
- Audio recordings, voice samples and music files you upload or record
- PDFs, documents and other files you attach as inputs or references
- Text you enter: prompts, instructions, chat messages and editing directions
- Conversation context: the relevant portion of your current session and summaries of prior turns needed for the model to understand your request
- Previously generated outputs that you feed back for further editing or extension
- Technical processing metadata: file format, resolution, duration and job parameters required to run the model
What is never sent to an AI model
We do not send your name, email address, phone number, payment or billing information, contacts, precise location, device advertising identifiers, or your connected social media account credentials to AI model providers. Requests are transmitted under ZorgSocial's own service credentials, so providers do not receive your ZorgSocial account identity. We do not use facial data to identify individuals and we do not build facial recognition databases.
4. The AI Models We Use
ZorgSocial works with a broad catalogue of third-party AI models so that the Easy Zorg agent can select the best model for each task β quality, speed, cost and capability all vary by model. The inventory below lists every model currently available through our platform, the provider behind it, the category of work it performs, and a link to that provider's own privacy policy so you can read it yourself.
The current inventory contains 529 models: 56 image generation models, 34 video generation models, 26 audio and text-to-speech models, and 413 large language models. You can search and filter the list below.
Many of these models are accessed through OpenRouter, a model routing and hosting layer. Where a model is routed through OpenRouter, OpenRouter's own privacy policy applies in addition to the model vendor's policy. Some models are open-weights community models with no vendor privacy policy of their own; for those, the policy of the hosting provider that runs the model applies, and we link to OpenRouter's policy as the routing layer.
Showing 40 of 529 models
Nano Banana
Image generation- Provider
- What it does
- Google text-to-image/edit model (alias of Gemini 2.5 Flash Image).
Nano Banana Pro
Image generation- Provider
- What it does
- Google high-quality text-to-image model (alias of Gemini 3 Pro Image).
GPT Image
Image generation- Provider
- OpenAI
- What it does
- OpenAI text-to-image and image-editing model (alias).
Seedream
Image generation- Provider
- ByteDance / BytePlus
- What it does
- ByteDance text-to-image generation model (alias).
Flux Pro
Image generation- Provider
- Black Forest Labs
- What it does
- Black Forest Labs FLUX text-to-image model (alias).
Black Forest Labs: FLUX.2 Flex
Image generation- Provider
- Black Forest Labs
- What it does
- Flexible FLUX.2 text-to-image model with tunable speed/quality.
Black Forest Labs: FLUX.2 Klein 4B
Image generation- Provider
- Black Forest Labs
- What it does
- Small 4B-parameter FLUX.2 image model for fast, cheap generation.
Black Forest Labs: FLUX.2 Max
Image generation- Provider
- Black Forest Labs
- What it does
- Highest-quality FLUX.2 text-to-image model.
Black Forest Labs: FLUX.2 Pro
Image generation- Provider
- Black Forest Labs
- What it does
- Production FLUX.2 text-to-image model balancing quality and speed.
ByteDance Seed: Seedream 4.5
Image generation- Provider
- ByteDance Seed
- What it does
- Text-to-image generation and editing.
ByteDance Seed: Seedream 5.0 Lite
Image generation- Provider
- ByteDance Seed
- What it does
- Lightweight, faster Seedream 5.0 image model.
ByteDance Seed: Seedream 5.0 Pro
Image generation- Provider
- ByteDance Seed
- What it does
- Flagship Seedream 5.0 image generation/editing model.
Google: Nano Banana (Gemini 2.5 Flash Image)
Image generation- Provider
- What it does
- Fast image generation and conversational image editing.
Google: Nano Banana 2 (Gemini 3.1 Flash Image Preview)
Image generation- Provider
- What it does
- Preview of next-gen fast image generation/editing.
Google: Nano Banana 2 (Gemini 3.1 Flash Image)
Image generation- Provider
- What it does
- Fast image generation and editing.
Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image)
Image generation- Provider
- What it does
- Lowest-cost, fastest Nano Banana image model.
Google: Nano Banana Pro (Gemini 3 Pro Image Preview)
Image generation- Provider
- What it does
- High-fidelity image generation with strong text rendering.
Krea: Krea 2 Large
Image generation- Provider
- Krea
- What it does
- Large text-to-image model with photorealistic output.
Krea: Krea 2 Medium Turbo
Image generation- Provider
- Krea
- What it does
- Faster, lower-latency Krea 2 Medium.
Meta: Muse Image
Image generation- Provider
- Meta
- What it does
- Meta text-to-image generation model.
Microsoft AI: MAI-Image-2.5
Image generation- Provider
- Microsoft AI
- What it does
- Microsoft text-to-image generation model.
Microsoft AI: MAI-Image-2.5 Pro
Image generation- Provider
- Microsoft AI
- What it does
- Higher-quality MAI-Image-2.5.
Microsoft AI: MAI-Image-2.6
Image generation- Provider
- Microsoft AI
- What it does
- Microsoft text-to-image generation model.
Microsoft AI: MAI-Image-2.6 Flash
Image generation- Provider
- Microsoft AI
- What it does
- Fast, low-cost MAI-Image-2.6.
OpenAI: GPT Image 1
Image generation- Provider
- OpenAI
- What it does
- Text-to-image generation and editing.
OpenAI: GPT Image 1 Mini
Image generation- Provider
- OpenAI
- What it does
- Cheaper, faster GPT Image 1.
OpenAI: GPT Image 2
Image generation- Provider
- OpenAI
- What it does
- Text-to-image generation and editing.
OpenAI: GPT Image 2.5 Flare
Image generation- Provider
- OpenAI
- What it does
- GPT Image 2.5 variant for image generation/editing.
OpenAI: GPT Image 2.5 Sunburst
Image generation- Provider
- OpenAI
- What it does
- GPT Image 2.5 variant for image generation/editing.
OpenAI: GPT-5 Image
Image generation- Provider
- OpenAI
- What it does
- GPT-5-based image generation and editing.
OpenAI: GPT-5 Image Mini
Image generation- Provider
- OpenAI
- What it does
- Cheaper, faster GPT-5 Image.
OpenAI: GPT-5.4 Image 2
Image generation- Provider
- OpenAI
- What it does
- GPT-5.4-based image generation and editing.
Qwen: Qwen Image 3
Image generation- Provider
- Qwen
- What it does
- Text-to-image generation with strong text rendering.
Qwen: Qwen Image 3 Pro
Image generation- Provider
- Qwen
- What it does
- Higher-quality Qwen Image 3.
Recraft: Recraft V3
Image generation- Provider
- Recraft
- What it does
- Design-oriented text-to-image model.
Recraft: Recraft V4
Image generation- Provider
- Recraft
- What it does
- Design-oriented text-to-image model.
Recraft: Recraft V4 Pro
Image generation- Provider
- Recraft
- What it does
- Higher-quality Recraft V4.
Recraft: Recraft V4 Pro Vector
Image generation- Provider
- Recraft
- What it does
- Generates editable vector (SVG) graphics.
Recraft: Recraft V4 Styles
Image generation- Provider
- Recraft
- What it does
- Image generation with custom brand/style presets.
| AI Model | Provider | Category | What it does | Provider Privacy Policy |
|---|---|---|---|---|
| Nano Banana | Image generation | Google text-to-image/edit model (alias of Gemini 2.5 Flash Image). | Read policy | |
| Nano Banana Pro | Image generation | Google high-quality text-to-image model (alias of Gemini 3 Pro Image). | Read policy | |
| GPT Image | OpenAI | Image generation | OpenAI text-to-image and image-editing model (alias). | Read policy |
| Seedream | ByteDance / BytePlus | Image generation | ByteDance text-to-image generation model (alias). | Read policy |
| Flux Pro | Black Forest Labs | Image generation | Black Forest Labs FLUX text-to-image model (alias). | Read policy |
| Black Forest Labs: FLUX.2 Flex | Black Forest Labs | Image generation | Flexible FLUX.2 text-to-image model with tunable speed/quality. | Read policy |
| Black Forest Labs: FLUX.2 Klein 4B | Black Forest Labs | Image generation | Small 4B-parameter FLUX.2 image model for fast, cheap generation. | Read policy |
| Black Forest Labs: FLUX.2 Max | Black Forest Labs | Image generation | Highest-quality FLUX.2 text-to-image model. | Read policy |
| Black Forest Labs: FLUX.2 Pro | Black Forest Labs | Image generation | Production FLUX.2 text-to-image model balancing quality and speed. | Read policy |
| ByteDance Seed: Seedream 4.5 | ByteDance Seed | Image generation | Text-to-image generation and editing. | Read policy |
| ByteDance Seed: Seedream 5.0 Lite | ByteDance Seed | Image generation | Lightweight, faster Seedream 5.0 image model. | Read policy |
| ByteDance Seed: Seedream 5.0 Pro | ByteDance Seed | Image generation | Flagship Seedream 5.0 image generation/editing model. | Read policy |
| Google: Nano Banana (Gemini 2.5 Flash Image) | Image generation | Fast image generation and conversational image editing. | Read policy | |
| Google: Nano Banana 2 (Gemini 3.1 Flash Image Preview) | Image generation | Preview of next-gen fast image generation/editing. | Read policy | |
| Google: Nano Banana 2 (Gemini 3.1 Flash Image) | Image generation | Fast image generation and editing. | Read policy | |
| Google: Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) | Image generation | Lowest-cost, fastest Nano Banana image model. | Read policy | |
| Google: Nano Banana Pro (Gemini 3 Pro Image Preview) | Image generation | High-fidelity image generation with strong text rendering. | Read policy | |
| Krea: Krea 2 Large | Krea | Image generation | Large text-to-image model with photorealistic output. | Read policy |
| Krea: Krea 2 Medium | Krea | Image generation | Mid-size text-to-image model. | Read policy |
| Krea: Krea 2 Medium Turbo | Krea | Image generation | Faster, lower-latency Krea 2 Medium. | Read policy |
| Meta: Muse Image | Meta | Image generation | Meta text-to-image generation model. | Read policy |
| Microsoft AI: MAI-Image-2.5 | Microsoft AI | Image generation | Microsoft text-to-image generation model. | Read policy |
| Microsoft AI: MAI-Image-2.5 Pro | Microsoft AI | Image generation | Higher-quality MAI-Image-2.5. | Read policy |
| Microsoft AI: MAI-Image-2.6 | Microsoft AI | Image generation | Microsoft text-to-image generation model. | Read policy |
| Microsoft AI: MAI-Image-2.6 Flash | Microsoft AI | Image generation | Fast, low-cost MAI-Image-2.6. | Read policy |
| OpenAI: GPT Image 1 | OpenAI | Image generation | Text-to-image generation and editing. | Read policy |
| OpenAI: GPT Image 1 Mini | OpenAI | Image generation | Cheaper, faster GPT Image 1. | Read policy |
| OpenAI: GPT Image 2 | OpenAI | Image generation | Text-to-image generation and editing. | Read policy |
| OpenAI: GPT Image 2.5 Flare | OpenAI | Image generation | GPT Image 2.5 variant for image generation/editing. | Read policy |
| OpenAI: GPT Image 2.5 Sunburst | OpenAI | Image generation | GPT Image 2.5 variant for image generation/editing. | Read policy |
| OpenAI: GPT-5 Image | OpenAI | Image generation | GPT-5-based image generation and editing. | Read policy |
| OpenAI: GPT-5 Image Mini | OpenAI | Image generation | Cheaper, faster GPT-5 Image. | Read policy |
| OpenAI: GPT-5.4 Image 2 | OpenAI | Image generation | GPT-5.4-based image generation and editing. | Read policy |
| Qwen: Qwen Image 3 | Qwen | Image generation | Text-to-image generation with strong text rendering. | Read policy |
| Qwen: Qwen Image 3 Pro | Qwen | Image generation | Higher-quality Qwen Image 3. | Read policy |
| Recraft: Recraft V3 | Recraft | Image generation | Design-oriented text-to-image model. | Read policy |
| Recraft: Recraft V4 | Recraft | Image generation | Design-oriented text-to-image model. | Read policy |
| Recraft: Recraft V4 Pro | Recraft | Image generation | Higher-quality Recraft V4. | Read policy |
| Recraft: Recraft V4 Pro Vector | Recraft | Image generation | Generates editable vector (SVG) graphics. | Read policy |
| Recraft: Recraft V4 Styles | Recraft | Image generation | Image generation with custom brand/style presets. | Read policy |
We update this inventory whenever we add or remove a model. Material changes to the categories of AI providers we use are notified in the app and reflected in the "Last Updated" date above. If you would like to know which specific model processed a particular job, you can see it in that job's details in the app, or contact legal@zorgsocial.com.
5. Consent: Explicit, Informed and Withdrawable
We rely on your explicit consent as the legal basis for transmitting your content to third-party AI models. Consent works as follows:
- First use: the first time you use an AI Feature in the mobile app, we display a consent screen that identifies the categories of data to be sent and the AI providers involved. No content is transmitted until you affirmatively accept.
- Just-in-time approval: every meaningful generation, edit or upload action shows what will be processed and, where known, which model will process it, and requires your approval before it runs.
- Granular control: you can consent to some AI Features and decline others. Declining AI processing does not prevent you from using the non-AI parts of the Service, such as scheduling, publishing, calendars and analytics.
- Third-party rights: if you upload photographs or recordings of identifiable individuals, you confirm that you have the necessary rights and, where required by law, their consent. You must not upload biometric or special-category data of others without a lawful basis.
- No dark patterns: consent is never bundled with unrelated terms, never pre-ticked, and never inferred from continued use of the app.
Withdrawing consent
You can withdraw your consent at any time in the app under Settings β Privacy β Third-Party AI Processing, or by emailing legal@zorgsocial.com. After withdrawal, no further content is transmitted to AI model providers. Withdrawal does not affect the lawfulness of processing carried out before withdrawal, and it does not delete content you previously generated β you can delete that separately at any time.
6. Why Your Content Is Processed
Your content is sent to AI models solely to deliver the feature you asked for. Specifically, to:
- Generate, edit, composite, retouch, upscale, extend or transform the images, videos, audio and text you request
- Analyse and describe reference media you attach so that outputs match your intent
- Enable the Easy Zorg agent to reason about your request, plan multi-step workflows and select appropriate tools and models
- Transcribe, translate, caption or dub audio and video at your direction
- Produce voiceovers, music and sound design from the briefs you provide
- Provide quality assurance, abuse prevention and safety filtering required by law and by our model providers' terms
We do not send your content to AI models for advertising, behavioural profiling, credit or employment decisioning, or any purpose other than fulfilling the specific request you make. We do not sell your content, and we do not permit model providers to use your content to train their general-purpose AI models under the enterprise and API terms in place with each provider.
8. How We Protect Your Content
We apply technical and organisational measures designed to protect your content throughout the AI workflow:
- Encryption in transit using TLS for all uploads, model requests and downloads
- Encryption at rest for stored media, prompts and generated outputs
- Access controls: only a limited number of authorised personnel can access production systems, under least-privilege and audit logging
- Provider due diligence: we assess each AI provider's security posture and contractual data-protection commitments before enabling it
- No-training commitments: we contractually require providers not to train their general-purpose models on your content
- Transient processing: providers retain your content only as long as needed to process the request and for limited abuse-monitoring periods defined in their API terms
- Deletion tooling: in-app controls to delete individual jobs, sessions, assets or your entire account
9. How Long We Keep AI Content
Your uploads, prompts and generated outputs are stored on ZorgSocial's own infrastructure so that you can access your job history, reuse sessions and re-download assets. We keep them for as long as your account is active and as needed to provide the Service.
You may delete individual jobs, sessions or assets at any time, and you may delete your account entirely. When you delete content, it is removed from active systems promptly and from backups within our standard backup rotation period. We may retain limited records where required by law, for tax and accounting purposes, or to resolve disputes β and AI model providers retain content only transiently as described in Section 8.
10. Your Rights Over AI-Processed Data
Subject to applicable law, including the UAE Personal Data Protection Law and, where it applies to you, the GDPR, you have the right to:
- Access the personal data we hold about you, including which AI Features processed your content
- Correct inaccurate personal data
- Delete your personal data and generated content
- Restrict or object to processing, including objecting to third-party AI processing
- Withdraw consent at any time, without affecting prior lawful processing
- Data portability: receive your data in a structured, commonly used, machine-readable format
- Lodge a complaint with your local supervisory authority
- Not be subject to a decision based solely on automated processing that produces legal or similarly significant effects β ZorgSocial does not make such decisions about you
To exercise any of these rights, email legal@zorgsocial.com or use the in-app privacy controls. We respond within 30 days, or sooner where required by law. You will not be charged, and exercising your rights will never degrade your access to the Service.
11. Children and Age Limits
The AI Features are not directed to children. You must be at least 18 years old, or the age of legal majority in your jurisdiction, to create a ZorgSocial account and to use AI Features. We do not knowingly collect personal data from children or process children's images through AI Features.
If you believe a child's data has been processed through our AI Features, contact legal@zorgsocial.com and we will investigate and delete the data promptly.
12. International Data Transfers
ZorgSocial is operated from the United Arab Emirates, and many of the AI model providers we use are located in the United States, the European Union, Singapore and China. When you use an AI Feature, your content may therefore be transferred to and processed in a country other than your own.
We only transfer personal data where a lawful transfer mechanism is in place β such as standard contractual clauses, an adequacy decision, or your explicit consent to the transfer for the specific AI Feature you requested. The location of each provider is shown in the AI Sub-Processors & Providers list, and the provider's own privacy policy explains its transfer practices.
13. Changes to this Policy
We may update this policy to reflect new AI Features, new model providers, or changes in law. When we make material changes β for example, adding a new category of AI provider or a new use of your content β we will notify you in the app and, where required, ask for your consent again before the change applies to your content.
The "Last Updated" date at the top of this page always reflects the current version. Continued use of an AI Feature after a non-material update constitutes acceptance of the updated policy.
14. Contact Us
If you have questions about how AI processes your personal data, or want to exercise your rights, contact us:
- Legal: legal@zorgsocial.com
- Product support: support@zorgsocial.com
ZorgSocial is operated by QUANTIMEDX COMPUTER SYSTEMS TRADING L.L.C, registered in Dubai, United Arab Emirates.