Overview

This FAQ gathers the most common questions customers ask about Akeneo's AI capabilities in the PIM, from creating configurations and managing prompts to understanding performance, limits, and data security.

Use this page to quickly find practical answers before contacting support or your Customer Success Manager.

If you're looking for detailed guides on creating or running AI Configurations, visit the related pages in this section:

• Creating and Managing AI Configurations
• AI Prompt Optimizer and Rewriter
• AI Configuration Security and Data Privacy

1. General usage

What is an AI Configuration in Akeneo PIM? An AI Configuration in Akeneo PIM defines how the AI generates, translates, or rephrases product data. It contains the prompt (the instruction for the AI), the target Attribute where results are stored, and the source Attributes that provide context. For a full introduction, see AI Configurations Overview.

Where do I find the AI Configurations page? Go to Settings → AI Configurations. From there, you can view, create, and manage all AI Configurations for your environment.

Who can create or edit AI Configurations? Only users with the appropriate Settings and AI Features permissions can create or edit AI Configurations. Permissions are managed by an administrator.

Can multiple users work on AI Configurations at the same time? Yes, but each AI Configuration should be edited by one person at a time to avoid overwriting. Akeneo PIM automatically warns you if another user is editing the same configuration.

What types of AI Configurations can I create? You can create configurations for content generation, translation, and asset auto-tagging. Generation and translation configurations apply to product Attributes; auto-tagging configurations apply to Asset Families and control how the AI generates tags for asset images.

2. Prompts and optimization

What makes a good prompt for an AI Configuration? A good prompt is clear, structured, and relevant. It should define:

• The role of the AI (e.g., SEO copywriter, translator, data analyst)
• The task to perform (e.g., generate, rewrite, translate)
• The output format and tone
• The data sources to consider (specific Attributes or assets)

For best results, use the C.R.A.F.T.+R framework and test your prompt using the Preview feature before saving.

What does "Relevance" mean in the C.R.A.F.T.+R framework? Relevance checks whether your prompt makes sense for the selected Attribute type. For example, asking the AI to "write a detailed paragraph" for a Number Attribute would score low on relevance.

Can I rewrite a prompt automatically? Yes. Click Rewrite Prompt in the AI Configuration editor. A side-by-side comparison shows your original prompt and the suggested optimized version. You can accept or keep your original. Accepting replaces your prompt in the text area but does not save automatically — you must click Save to confirm.

Does the Prompt Optimizer and Rewriter work for translation configurations? No. The AI Prompt Optimizer and Rewriter is designed for generation prompts only. It does not apply to translation-only AI Configurations.

Can I add channel-specific instructions to a prompt? Yes. In addition to the base prompt in an AI Configuration, you can add channel-specific generation instructions directly on each Channel. Open the Channel, go to the Generative AI tab, and fill in the prompt field. Channel-level instructions are appended to the base AI Configuration prompt at generation time. This lets you enforce destination-specific constraints (format, length, required sections) without duplicating AI Configurations. See Creating and Managing AI Configurations for details.

3. Execution and automation

How can I test my AI Configuration before automating it? Use the Preview panel in the AI Configuration editor. Select a product, Channel, and Locale, and click Preview to see the AI output before saving or deploying.

How do I generate content manually using an AI Configuration? In the Product Edit Form, click Ask AI and choose Smart Fill (for generation) or Smart Translate (for translation). You can run either option without a configuration for a quick one-off result, or select a specific AI Configuration from the dropdown for more control over tone, guidelines, or Reference Entity context.

How do I automate AI generation at scale? Use the Rules Engine. Go to Settings → Rules, create or open a rule, define conditions that identify which products to affect, and under Actions, select Generate attribute value with AI. Choose the desired AI Configuration from the list, then save and enable the rule. When the rule runs, the AI Configuration generates or updates values for all matching products automatically.

Will AI rules run continuously? No. Rules run according to your defined frequency or schedule (for example, every few hours). Akeneo PIM executes them as background jobs.

Will the Rules Engine override existing attribute values? Yes, if the rule has no condition filtering out products that already have a value. Apply appropriate filters in the rule conditions to avoid overwriting existing content.

4. Limits and performance

Is there a limit to how many AI Configurations I can create? Yes. Each environment supports up to 500 AI Configurations. If you need more, contact your Customer Success Manager.

How many products can AI process at once? The number depends on your data and assets. As a general reference, processing 20,000 products with 5 assets each may take up to 24 hours. Schedule large-scale automation during off-peak hours.

Can I run several AI rules simultaneously? You can, but staggering them optimizes performance. Running too many complex AI rules at the same time may delay processing.

What happens if my product has too many assets? Only the first 5 images (up to 32 MB total) and the first 2 PDFs (up to 20 MB total, first 20 pages each) are analyzed. Any additional assets are ignored. See AI Configuration Considerations & Limitations for the full list of asset constraints.

Are there any hard usage limits on AI calls? There are currently no hard per-customer usage limits, but usage is monitored internally. Future releases may introduce customer-specific fair usage thresholds. Contact your Akeneo representative if you run consistent large-scale automation.

5. Supported data and Attribute types

Which Attribute types can be generated using an AI Configuration? In Akeneo PIM, the following Attribute types can be used as output fields in an AI Configuration:

• Text
• Text Area
• Simple Select / Multi Select
• Number and Measurement
• Yes/No (Boolean)
• Reference Entity Single Link / Multiple Links
• Date

Can AI update Reference Entities directly? No. AI Configurations can fill Attributes that link to Reference Entities, but cannot modify the Reference Entity records themselves.

Can AI use Reference Entity data as context for generation? Yes. When a product is linked to a Reference Entity record (such as a Brand or Collection), Akeneo PIM automatically includes supported Attribute values from that record in the generation context. Supported types are Text Area, Simple Select, Multi Select, and Number. Images, Assets, and nested Reference Entity links are not included.

Can AI use category labels or hierarchy data? Not currently. Category data is not accessible to the AI during generation.

Can AI process assets stored outside of Akeneo? No. The AI can only analyze assets available within the Akeneo Asset Manager or linked Asset collections. The asset collection must include a main media Attribute for AI processing.

6. Auto-tagging

What is auto-tagging in Akeneo PIM? Auto-tagging is an AI feature that automatically generates tags for asset images when they are uploaded to an Asset Family. Akeneo PIM's AI analyzes the image content and applies relevant tags, saving time and maintaining consistency across your asset library.

Do I need an AI Configuration for auto-tagging to work? No. Auto-tagging works with default behavior even without an AI Configuration. An Asset Auto-Tagging AI Configuration is optional and lets you customize which tags are included or excluded (for example, excluding brand names or focusing on materials and colors).

How do I enable auto-tagging for an Asset Family? Go to Settings → Asset families, open the Asset Family, select the Properties tab, toggle on Enable auto-tagging generation, and optionally select an AI Configuration. Click Save. For a full guide, see AI Configurations for Auto-Tagging.

Does auto-tagging consume AI credits? Yes. Tagging operations consume AI credits. If your catalog reaches its credit limit, auto-tagging is paused until credits are replenished.

7. Data privacy and security

Is my data used to train AI models? No. Akeneo uses OpenAI Enterprise, which ensures zero data retention and no use of your data for AI model training.

What information is shared with the AI provider? Only the data required to generate the response: selected Attribute values, prompt text, Locale, and (if applicable) asset URLs. No customer-identifying information is included unless you add it manually to the prompt.

Can I restrict which users can access AI features? Yes. Administrators can use role-based permissions to control who can view, edit, or execute AI Configurations.

Does Akeneo store my prompts or AI results? Prompts and generated values are stored only in your PIM if you choose to save them. AI processing itself is stateless and ephemeral — data is not retained by the AI provider after each request.

For a full explanation of data flows and security practices, see AI Configuration Security and Data Privacy.

8. Troubleshooting common issues

9. Support and feedback

Who can I contact if I encounter problems? Reach out to your Customer Success Manager or contact Akeneo Support through the Help Center. Include the configuration name, rule ID, and timestamp of the issue to speed up investigation.

How can I share feedback about AI performance or usability? Akeneo welcomes customer feedback to improve the AI experience. You can submit feedback through your Customer Success contact or via in-app surveys when available.

Will Akeneo release more AI features? Yes. The AI roadmap continues to expand, with planned improvements to asset analysis, automation, and enrichment capabilities. Contact your Customer Success Manager for the latest roadmap updates.

Summary

Akeneo's AI Configurations combine automation, clarity, and control in the PIM. By understanding how AI Configurations work — and keeping these common questions in mind — you can confidently generate, translate, and optimize product data at scale while maintaining data integrity and brand consistency.

For further reading:

AI Prompt Optimizer and Rewriter — Learn how to improve prompt quality.

AI Configuration Security and Data Privacy — Review how Akeneo keeps your data protected.

Creating and Managing AI Configurations — Start building your first AI Configuration.