With artificial intelligence companies exhausting publicly available content to train their models, new opportunities are emerging for content creators to monetise their unused footage.
AI firms are offering YouTubers and other creators the chance to sell their surplus content for thousands of pounds, even as platforms enforce stricter rules on “synthetic” content in their databases.
How Much Does Unused Footage Fetch?
According to a report from Bloomberg, AI companies spend between £0.82 ($1) and £3.27 ($4) per minute on video footage, with prices varying based on quality and format.
Meanwhile, high-resolution content, such as 4K videos, tends to command higher rates than specialised footage, including drone recordings or 3D animations. Most videos, mainly unused content initially produced for platforms like YouTube, Instagram, or TikTok, typically sell in the £0.82 ($1) to £1.63 ($2) per minute range.
Content creators record hundreds of hours of video annually while producing content for platforms like YouTube, Instagram, and TikTok. However, a significant portion of that footage never gets published online, leaving it without a chance to generate revenue.
By selling this unused material to AI companies, creators can unlock a new income stream, providing an alternative to traditional brand sponsorships and advertising deals.
Who Is Paying For These Footage?
Sources familiar with the matter said that several AI companies, including OpenAI, Google, and AI-focused media firm Moonvalley, are compensating hundreds of content creators for access to their unpublished video footage. Due to its uniqueness, this exclusive content, which hasn’t been shared online, is highly sought after for training artificial intelligence systems.
It is worth noting that multiple companies like OpenAI and Adobe have introduced AI tools capable of generating realistic videos from text descriptions. However, developing such technology requires enormous datasets—often amounting to millions of hours of video content—for practical training.
‘At Moonvalley, the vast majority of our training data is coming sourced directly from content creators & filmmakers who are interested in licensing any video content they own,’ Moonvalley told Bloomberg.
An Alternative To Content Scraping?
Selling unused footage to companies may seem like a safer alternative for AI companies to train their data. In the past, companies have used content scraping to use publicly available data to train their AI models. While many have started out using public domain content and those whose trademark and copyright have expired, many have resorted to scraping copyrighted content and even content from online users without their consent.
In 2024, leaked documents from AI company Runway revealed a massive database of YouTube channels that the company scrapped data from to use their training model. Moreover, it has also been alleged that the company scrapped data from pirated copies of anime series and Internet Archive pages.
It is worth noting that Google had previously cautioned OpenAI against utilising videos from its platform to train the AI model Sora, emphasising that such actions would violate the platform’s Terms of Service.
‘From a creator’s perspective, when they upload their hard work to our platform, they have certain expectations. One of those expectations is that the terms of service must be abided by. It does not allow for things like transcripts or video bits to be downloaded, which clearly violates our terms of service. Those are the rules of the road regarding content on our platform,’ YouTube CEO Neil Mohan told Bloomberg in April 2024.
For context, YouTube has stringent measures on so-called “synthetic” content. Creators are required to disclose when they’ve created realistically altered or synthetic content, including using AI tools.
Mutually Beneficial in the Longer Run
The practice of YouTubers selling their unused footage to AI companies represents a mutually beneficial opportunity. For creators, it offers a new revenue stream by monetising content that would otherwise remain dormant.
For AI companies, this untapped video material provides a valuable resource for training their models, enabling advancements in artificial intelligence capabilities. As the demand for high-quality, diverse datasets grows, this collaboration highlights how creators and tech innovators can work together to fuel the next generation of AI development.