AI Quick Edit is Canto's in-app generative AI image editing feature. It allows users to perform background replacement, smart cropping, resizing, text overlays, and prompt-based edits without leaving the DAM. For more detail on what the feature does and how to use it, see the AI Quick Edit article.
Quick answers
Most commonly asked questions about AI Quick Edit:
| Question | Answer |
|---|---|
| Which AI service processes my images? | Google Gemini (via Google Cloud / Vertex AI). Google is an optional sub-processor — it only processes data when AI Quick Edit is enabled on your account. |
| Is my content stored by Google? | No. Data submitted via the Gemini API is processed transiently to generate the output and is not stored by Google beyond the API call. |
| Is my content used to train AI models? | No. Google's API commitments confirm customer data submitted via the Gemini API is not used to train or fine-tune Google's models. Canto's enterprise agreement with Google contractually prohibits this. |
| Who owns the AI-generated output? | You do. Customers retain ownership of their data and all AI-generated outputs produced from it. |
Which AI service processes your images
When a user submits an image and prompt to AI Quick Edit, both are sent to Google Gemini via Google Cloud / Vertex AI.
Google is an optional sub-processor under Canto's Data Processing Agreement. It processes customer data only when AI Quick Edit is enabled on the account and actively used by a user. If AI Quick Edit has not been enabled for your tenant, no data is transmitted to Google.
Canto's AWS infrastructure handles storage and delivery of your assets before and after processing. The image modification itself is performed by Google Gemini.
Please note: Google Gemini processing may occur outside the EU/EEA. EU/EEA customers should be aware that assets submitted to AI Quick Edit may be processed by Google infrastructure in the US or another non-EEA jurisdiction. Cross-border transfers are covered by Google's Cloud Data Processing Addendum, which includes Standard Contractual Clauses. |
What data is sent to Google Gemini
When AI Quick Edit is used, the following is transmitted to the Gemini API:
- The full image asset as submitted, including any embedded EXIF metadata or image content.
- The full text prompt entered by the user, without modification or filtering by Canto.
The image and prompt are passed to the API as-is. If an asset contains personal data — for example, images of identifiable individuals, documents with personal information, or metadata with names — that personal data will be included in the content transmitted to Google.
Please note: As the data controller, your organisation is responsible for ensuring you have an appropriate lawful basis for submitting personal data to AI Quick Edit. Avoid including personal data in prompts where it is not necessary. AI Quick Edit is not designed for processing special category data (health data, biometric data, racial or ethnic origin, and similar). If you have assets containing special category data, review your legal basis and obligations as controller before submitting them. |
Model training
Customer data submitted to AI Quick Edit is not used to train AI models — neither Canto's nor Google's.
- Canto does not train foundational or shared AI models using customer production data.
- Google's published commitments for the Gemini API (via Google Cloud / Vertex AI) confirm that data submitted by enterprise API customers is not used to train or fine-tune Google's models without explicit customer permission.
This applies to both the image assets and the text prompts submitted. Neither is retained or repurposed by Google for model improvement.
Data storage and deletion
Your original assets remain stored in Canto's AWS infrastructure for the duration of your contract. AI Quick Edit does not create a separate copy of assets in Google's systems — data is processed in transit and Google does not store submitted content beyond the API call.
Within Canto, standard data retention rules apply:
- Assets deleted by users are held in Trash for 30 days before permanent deletion.
- Backups are retained for 30 days following deletion or contract termination, after which they are permanently deleted.
- Upon contract termination, all customer data is removed from Canto's systems 30 days after the contract end date.
Content ownership
Customers retain full ownership of their data, including any AI-generated outputs produced by AI Quick Edit. Canto does not claim any intellectual property rights over modified assets.
The AI-modified version of an asset is saved as a new version or new asset within your Canto library, in the same tenant environment as the original. It is subject to the same access controls, retention settings, and organisational policies as any other asset.
Please note: Responsibility for ensuring that AI-generated outputs comply with applicable copyright, IP, or brand guidelines rests with your organisation as the data controller and content owner. Canto recommends reviewing AI-generated outputs before publication or distribution. |
Frequently asked questions
Can I disable AI Quick Edit for my organisation?
Yes. AI Quick Edit must be explicitly enabled by your Canto Account Manager. To disable it after enabling, contact your Account Manager or submit a request via the Help Center. Disabling the feature at the tenant level stops all data transmission to Google and prevents all users in that tenant from accessing it.
What happens if a user submits an image containing personal data?
The image, including any personal data it contains, will be transmitted to Google Gemini for processing. Canto does not screen or filter image content before submission. As the data controller, your organisation is responsible for ensuring that submitting such assets is consistent with your data protection obligations and any applicable instructions from data subjects.
Is Google involved in any other Canto AI features?
No. Google is an optional sub-processor used exclusively for AI Quick Edit.
Does AI Quick Edit support bulk processing?
No. AI Quick Edit is a user-initiated, asset-by-asset feature. Each image must be submitted individually. Bulk or automated processing workflows are not supported.
Where is my image stored after it has been edited?
The AI-modified output is saved as a new version or new asset within your Canto library on AWS infrastructure in your designated hosting region. It is not stored in Google's systems after the API call completes.
Is AI Quick Edit compliant with the EU AI Act?
Yes. Canto has reviewed AI Quick Edit against the EU AI Act and confirmed it is categorised as a minimal risk AI system.