AI-generated Assets Marketplace

Anyone who has worked on product design, website development, or marketing campaigns for small businesses or early-stage startups understands the cost of creating new visual assets from scratch. That is why stock image platforms like Shutterstock and Getty Images provide a vast catalog of readily available assets for a reasonably cheap price.

It might surprise many that the stock visual asset market is a multi-billion dollar annual revenue business. However, with the advent of generative AI, there is an open window of opportunity for a new type of visual asset marketplace to emerge that serves creators using generative AI models to craft high-quality assets. Midjourney recently updated its Terms & Conditions, granting full ownership of the output to its paid subscribers and providing an excellent opportunity to target its users.

Assumptions

• Quality output prompts are very hard, so the target audience still prefers to buy existing assets that meet their requirement
• Thanks to the low cost of production, one could offer the most extensive catalog addressing very niche demands

Some additional features

• Require prompt submission along with the output asset to check for IP or artist-style infringement
• Use ZKP or simple hash of prompt to validate that the marketplace is running infringement checks without revealing prompts
• Unlike a prompt marketplace, the visual asset is the product being sold, so tailor the experience for assets browsing and purchase flow
• Beyond simple stock image T&C, enable more advanced licensing that allows creators to incorporate the assets into their IP and participate in the upside (narrative universe, animation, NFT projects, gaming)