Our policies and stance on AI

Summary: - Generally, we do not allow AI-generated content on our platform. - We do not endorse the usage of generative AI in connection with our platform. - If you are making a documentary about AI or a similar editorial project, you may qualify for an exception.

Halfspire is a platform for genuine expression of human creativity. Because of this, we place some restrictions on the usage of AI to create content for Halfspire.

When submitting content to Halfspire, you will be required to accurately specify whether or not generative AI was used.

We generally do not accept non-editorial submissions that use generative AI at all (with the exception of ethically trained text correction models, such as Grammarly and/or your device's built-in autocorrect feature - we do not classify these models as generative AI).

In exceptional cases, editorial submissions (such as documentaries) may use generative AI for demonstration purposes in connection with relevant commentary (created by a human). Additionally, the usage of generative AI in these submissions is strictly limited to depicting the output of AI to explain the topic of AI. Tutorials are not considered editorial submissions in this context, and we will decide what qualifies as an "exceptional case" at our own discretion.

The following official policy exists to help us moderate editorial/educational usage of AI, and to moderate/prevent its usage in marketing material. The official policy does not take effect in favor of AI usage if your submission is rejected, taken down, or restricted due to AI usage.

Official policy

We do not accept submissions that are primarily made using generative AI models that fall into these categories (meaning more than 60% of the total content is generated using any of these models):

• Text generation (such as GPT and Gemini models)

• Image generation (such as Midjourney, Stable Diffusion, and DALL-E) (additional policies apply, please see full document)

Ethically trained text generation models that only complete individual words or correct existing content (such as your device’s autocomplete or Grammarly) are not considered text generation models in this context. We may determine what constitutes acceptable use at our own discretion.

Additionally, we generally do not allow the usage of AI-generated images anywhere within content made for Halfspire. The only exception to this rule is editorial usage - such as using such images as examples in a documentary film about AI. Editorial usage will be evaluated on a case-by-case basis.

We also prohibit the usage of AI to generate metadata (such as titles, posters, descriptions, and cover images).

If you use our promotion tools to manage your ads (or if we manage ads for you), please note that ads using AI-generated images and/or AI-generated text will be rejected.

If you use AI-generated images and/or AI-generated text in an external ad (such as pay-per-click ads or billboards that are not managed via Halfspire), you may not use the Halfspire logo or brand assets in the ad. Halfspire Originals are not allowed to use AI-generated images and/or AI-generated text in any marketing material.

We understand that AI is constantly evolving, but we also believe that our current policies best reflect both our values and the best interest of our community. These policies exist to protect human creativity on our platform.