AI Policy

1. What AI does on the platform

AI is a working tool inside SAIMA, not a decision-maker. It drafts and transforms material that you review, approve and publish. Every AI feature on the platform produces a draft; none of them publishes, spends budget or contacts an audience on its own.

• Text generation — Marketing copy, captions, campaign ideas and content plans, produced from the brief and brand details you provide.
• Image generation — Visual assets created from your text prompts for use in posts and campaigns.
• Video generation — Short video assets generated from prompts and supplied material.
• Audio and voice — Voice-over and audio tracks synthesised from text you supply.
• Suggestions and analysis — Recommendations on scheduling, audience and campaign structure, derived from the data in your workspace and connected accounts.

2. Which AI providers we use

SAIMA does not train or host its own foundation models. Requests are sent to specialist providers that process them and return a result.

We work with a small number of vetted providers, each engaged for a specific function: text and visual generation, video generation, and voice synthesis. Every provider is bound by contract to process your material only to fulfil your request, and is prohibited from using it to train general-purpose models.

The current list of providers is maintained internally and available to enterprise customers on request. Because infrastructure and provider arrangements are subject to change, we publish the categories here and keep the named list in a document we can update without amending this policy. Material changes will be notified before they take effect.

3. Human review is required, not optional

AI output can be wrong. It can state facts that are not true, misattribute a quotation, produce a figure that looks plausible and is not, or generate an image that misrepresents a product. This is a property of the technology, not a defect in a specific request.

You must review every AI output before publishing it, sending it to an audience, or relying on it in a commercial decision. This applies with particular force to claims about price, availability, medical or financial matters, regulated products, and any statement about a third party.

The platform is built around this: AI features produce drafts into your workspace, and publishing is always a separate, explicit action taken by a person.

4. What we do not guarantee

We do not guarantee that AI output is accurate, complete, current, original, or free of third-party rights. We do not guarantee that the same prompt produces the same result twice — these models are probabilistic by design.

We do not guarantee that AI-generated content will perform commercially, reach a given audience, or comply with the advertising rules of any platform you publish to. Meeting the policies of Meta, Google, TikTok, Snapchat and any other destination remains your responsibility as the publisher.

5. Prohibited uses of AI features

You may not use SAIMA's AI features to produce material that impersonates a real person or organisation, fabricates a testimonial or review, generates deceptive claims about a product, creates content depicting a real individual without their consent, or produces anything unlawful under the laws of the Kingdom of Saudi Arabia.

You may not use the platform to generate content that infringes a third party's intellectual property, or to strip attribution from protected material. The Acceptable Use Policy sets out the full list and the consequences of breach.

6. AI credits

AI features consume credits from your subscription plan. Each plan includes a monthly allowance, and the allowance resets with your billing cycle.

Credits are consumed at the moment a generation request is processed by the provider, including when you discard the result. A request that fails because of a platform error does not consume credits.

7. Your data and the models

What you send to an AI feature — your prompt and any material attached to it — is transmitted to the provider that fulfils the request. We do not use your content to train general-purpose AI models, and we do not permit our providers to do so under our agreements with them.

The AI Data Use policy sets out in detail what is sent, to whom, where it is processed and how long it is kept.

8. Questions and reports

If an AI feature produces harmful, infringing or clearly incorrect output, report it to support@saimahub.ai with the prompt and the result. Reports of this kind directly inform our safeguards.