Privacy policy

Who we are

The controller of your personal data is [Company name], which operates the Tabaqat platform, at [National address], commercial registration [Commercial registration number].

For anything about your data and your rights, write to hello@tabaqat.app.

Your pictures and projects

Your device opens the pictures and processes the projects you make. Autosaves and recovery copies are kept in your browser's storage on that device, and they never reach us.

Clearing this site's data in your browser deletes them, so download project files you want to keep.

On-device AI tools

Select subject, remove background and upscale run inside your browser. The first time you use one, your browser downloads the model from Hugging Face, which receives that download request and your IP address. Your picture is not sent anywhere.

Generative tools (Pro)

When you use Edit with AI, we send your description and the part of the picture you are editing, with some of the picture around it, or the whole picture if nothing is selected, to Google's Gemini API, which returns the result. When you use Create picture, we send only your description. We pass the result straight back to your browser and do not store your picture, your description or the result.

Google processes these requests as our service provider under its terms for paid Gemini API services, which do not allow it to use them to improve its products or train its models. Google does keep requests and results for a limited period to detect abuse and for legally required disclosures, and may process them in any country where it has facilities. Pictures Google generates carry its invisible SynthID watermark.

When you use the design assistant, we send what you type, a description of your open document (its size, its layer names, the words in its text layers, their styles and positions, and what its QR codes contain) and, when the assistant needs to look at your design, a small picture of it, to Anthropic's Claude API, which answers with the steps your browser then carries out on your design. We store none of this, except what we keep when the assistant is suspended, as described next. Anthropic processes these requests as our service provider under its commercial terms, which do not allow it to train its models on them, and keeps them for a limited period for safety and abuse monitoring.

If you clearly and deliberately ask the design assistant for something its usage rules forbid (see the terms of use), the assistant is suspended for your account until we review the request. For that review we keep the account or access key, the rule that was broken, a one-sentence note the assistant writes about what was asked, the first 500 characters of your request, and when the suspension began and was lifted. We use these only to review the suspension and to keep the service within its rules, and delete them a year after the suspension is lifted. If the assistant is used for things other than design, we count how many times that happened that day and, from the second time, pause the assistant for an hour; we delete the count and the pause after two days. If the assistant receives verbal abuse (insults, swearing or threats at it or at others), we count it: the first two times are warnings, and the third suspends the assistant for your account until we review it as above; we delete that count after 30 days without new abuse, or when the suspension is lifted.

To enforce plan limits we keep, for each account or access key, a count of generative images per day and per month, of generative requests and declined requests per day, and of design assistant requests per day and per month, and of photo library searches and downloads per day, without any picture, description or text.

The free photo library

When the design assistant looks for a real photo for you, we send the search words it writes from your request (usually in English) to Openverse, the search engine of the WordPress Foundation, which returns pictures licensed CC0 or marked as public domain. We then fetch the thumbnails and the picture you choose from the sites that host them (such as Flickr and Wikimedia Commons) through our servers, so the host sees neither your IP address nor your design. We send Openverse no name, no design and no picture of yours.

We keep search results and thumbnails in Cloudflare's cache for at most a day, without linking them to your account. We do not store your search words with your account.

Payments

When subscriptions open, payments will be handled by Paddle, which sells the subscription as merchant of record and processes your payment details under its own privacy policy. We will receive your email address and subscription status from Paddle, never your card details.

The suggestion box

When you send a suggestion from the home page or the editor, we keep its text, your email address if you give one, the page and language you sent it from, and when you sent it, in our database at Cloudflare. We use these to read suggestions and improve the platform, and to write to you about your suggestion if you left your email address. We do not use your email address for anything else.

To stop spam, we limit how many suggestions can be sent from one address in a day. For that we keep a SHA-256 hash of your IP address and the date instead of the address itself. It is not linked to your suggestion, and we delete it the next day.

You can ask us to delete your suggestion at any time by writing to hello@tabaqat.app, with its text or the email address you sent it with, so we can find it.

Hosting and request logs

Cloudflare hosts the Tabaqat platform. To deliver the site and protect it from abuse, Cloudflare processes your IP address and the details of each request, and we can see request logs to fix problems.

Cookies and storage in your browser

We do not use advertising, analytics or tracking cookies. Cloudflare may set one strictly necessary cookie, __cf_bm, to tell automated traffic from people and protect the site. It expires within 30 minutes and is not used to track you.

The editor keeps your settings, such as language and panel layout, your autosaved projects, and an access key if you enter one, in your browser's local storage. Of these, only the access key ever reaches us, when you use a generative tool, to check your subscription.

How long we keep data

• Request logs at Cloudflare: deleted automatically within 7 days.
• Generative image usage counts: 12 months, then deleted.
• Your email address and subscription details: while your subscription is active, then for as long as applicable laws require, then deleted.
• Your messages to us: two years from our last exchange, then deleted.
• Suggestions sent through the suggestion box: two years from when you sent them, then deleted, or sooner if you ask.
• The IP address hash used to limit suggestions: deleted the next day.
• Pictures and descriptions sent to the generative tools: we keep none; Google keeps them for the limited period described above.
• Your messages to the design assistant and what is sent with them: we keep none, apart from what a suspension keeps (next item); Anthropic keeps them for the limited period described above.
• A design assistant suspension (the account, the rule, the note, the first 500 characters of the request and the times): until we review it, then for a year after it is lifted, then deleted.
• Counts of design assistant use for things other than design, and the pauses they cause: two days after the day they count, then deleted.

Transfers outside Saudi Arabia

Some of the service providers we rely on process data outside the Kingdom of Saudi Arabia:

• Google (United States and other countries): generative tools.
• Anthropic (United States): the design assistant.
• Openverse (United States): the free photo library search.
• Cloudflare (global network): hosting, protection, request logs, stored suggestions and design assistant suspensions.
• Paddle (United Kingdom): payments and subscriptions.
• Hugging Face (United States): AI model downloads to your browser.

We send each provider only the minimum it needs, rely on its contractual data protection commitments, and make these transfers as permitted by the Regulation on Personal Data Transfer outside the Kingdom.

Children

The Tabaqat platform is not directed at children, and the generative tools are only for people aged 18 and over. We do not knowingly collect personal data from anyone younger, and we delete any we learn of.

Your rights

The Personal Data Protection Law gives you the right to:

• Be told how and why your data is collected and processed, which this policy explains.
• Access the data we hold about you and get a copy of it in a clear format.
• Have your data corrected, completed or updated.
• Ask for your data to be destroyed when it is no longer needed.
• Withdraw your consent to processing at any time, where consent is its basis.

To use any of these rights, write to hello@tabaqat.app. We will reply within 30 days. If you are not satisfied with our reply, you can complain to the Saudi Data and AI Authority (SDAIA).

Changes to this policy

We will update the date at the top of this page when this policy changes, and announce significant changes on the site before they take effect.

If the Arabic text of this policy and this English translation differ, the Arabic text prevails.