Shutterstock is transforming the way AI corporations access training data through an revolutionary ‘research license’ approach, starting first with a creative AI technology company Lightricks. The partnership announced today enables Lightricks to train the LTXV model to generate open-source video using Shutterstock’s extensive HD and 4K video library.
The recent licensing model addresses a key challenge in the development of artificial intelligence: the high cost of access to high-quality training data. It allows corporations to start with a smaller research license to test and experiment before committing to costlier business licenses.
Increasing the accessibility of ethical AI development for startups
“Many companies and modeling trainers have gone the route of unauthorized data acquisition [instead of] we are making the necessary investments to achieve the quality and level of trust necessary to develop commercially viable models,” said Daniel Mandell, global head of data licensing and artificial intelligence at Shutterstock, in an exclusive interview with VentureBeat. “However, we do not believe that financial investments should be a barrier for those who want to enter this space with an ethical approach.”
This two-phase approach could change the way startups approach AI development. Craig Andrews, global PR manager at Lightricks, describes it as “an inflection point for smaller, more agile developers looking to explore innovative applications of generative AI without the high up-front costs of traditional licensing.”
Legal protection and fair remuneration in the era of artificial intelligence
The timing is significant and comes amid growing legal scrutiny of AI training data practices. Several major artificial intelligence corporations are facing lawsuits over the alleged unauthorized use of copyrighted materials to train models. Shutterstock’s approach offers a legitimate alternative while ensuring that content creators are compensated.
“We set the standard for ethical AI development while ensuring creators are fairly compensated for their work,” explains Andrews. “This approach not only increases trust in the creative ecosystem, but also establishes a sustainable framework for responsible AI innovation.”
Revenue sharing: advantages for AI creators and corporations
Shutterstock has implemented a revenue-sharing model in which contributors receive 20% of revenue from data licensing agreements. Creators also can opt out of getting their content used for AI training, although Mandell notes that only about 1% have chosen to do so.
Lightricks plans to use licensed video data to improve LTXV, an open-source video generation model released last month. The model has already gained significant popularity with “thousands of downloads”. GitHub AND Face Hugging– says Andrews. A notable use case is real-time video generation for interactive e-commerce.
The partnership goals to address technical issues related to AI video generation, particularly motion consistency in longer videos. “One of the biggest technical hurdles in AI video generation is achieving consistent motion and structure across longer video segments without sacrificing quality,” Andrews says. “Shutterstock’s high-quality video library provides a comprehensive dataset that helps us address this challenge.”
For Shutterstock, this partnership represents a strategic shift in its business model. The company has already partnered with major AI corporations, including Nvidia, MetaAND OpenAI. Mandell emphasizes that a research licensing model could democratize access to high-quality training data for smaller organizations and research institutions.
Setting recent industry standards for ethical AI development
The collaboration also reflects a growing trend toward transparency and ethical considerations in AI development. Lightricks has open-sourced LTXV to promote collaboration and innovation, and Shutterstock’s approach to licensing ensures content creators are properly compensated.
“The important message is that companies, regardless of size and funding, no longer have an excuse to use unlicensed content for training purposes,” concludes Mandell. “There is a better way to enter this growing market.”
This partnership could set a recent standard for AI corporations’ access to training data, potentially impacting industry practices as concerns grow about the sources of AI training data. The success of this model could determine whether other content providers follow Shutterstock’s lead in creating more flexible and accessible licensing options for AI development.