The Copyright Challenges in the Realm of Generative AI

News
September 22, 2023

Persistent Copyright Challenges in the World of Generative AI

Are you grappling with the persistent copyright challenges in the world of generative Artificial Intelligence (AI)? If so, you’re not alone. The copyright issues surrounding generative AI are not going away anytime soon. From lawsuits against companies like OpenAI to debates about fair use and opt-out options for artists, the landscape is complex and ever-evolving. In this article, we’ll explore the ongoing copyright challenges in the world of generative AI and discuss the importance of addressing these issues to protect the rights of artists and creators.

Generative AI has thrust a plethora of copyright concerns into the spotlight. Just this week, authors including George R.R. Martin, led by the Authors Guild, initiated a lawsuit against OpenAI, alleging that the company’s viral AI-powered chatbot, ChatGPT, was trained on their literary creations without their informed consent or knowledge.

However, OpenAI is not the sole entity grappling with these challenges. During Disrupt 2023, Anastasis Germanidis, one of the co-founders of Runway, a firm pioneering generative AI tools for the creative realm, disclosed that his company is in the process of “exploring” the most appropriate approach to training AI models using artists’ and creators’ content.

“We’re collaborating closely with artists to determine the most effective approaches,” Germanidis stated. “We’re actively exploring potential data partnerships as part of our efforts to expand and develop the next wave of models.”

Similar to numerous generative AI startups, Runway maintains confidentiality regarding the precise origins of its training data. This includes Gen-2, which specializes in generating videos from text prompts. In an academic paper providing insight into Gen-2’s architecture, Runway discloses that the model underwent training using an internal dataset comprising 240 million images and 6.4 million video clips. However, specific details about the data sources remain undisclosed.

The possibility exists that certain portions of this data could be subject to copyright protection. If indeed copyrighted material is included, it could potentially lead to legal complications for Runway in the future.

In recent times, artists have taken legal action against companies like Stability AI, Midjourney, and DeviantArt. These artists argue that the models created by these companies infringe upon their copyrights. This infringement arises from the models being trained on the artists’ original works and subsequently producing outputs that mimic their unique styles. Additionally, Getty Images has initiated legal proceedings against Stability AI, alleging that the company copied and processed millions of images, along with their associated metadata, which are owned by Getty Images in the U.K.

Certain companies in the generative AI sector contend that they fall under the protection of the fair use doctrine, particularly within the U.S. However, it’s a complex issue that is unlikely to see a definitive resolution in the near future.

In an effort to proactively protect themselves from potential legal disputes down the line, a select few generative AI providers, including Stability AI, have implemented mechanisms that allow artists to opt out of having their work used in model training. Just recently, OpenAI established a communication channel enabling artists to communicate their desire for their artwork not to be used in future model training. Additionally, some companies have initiated collaborative funds designed to share a portion of the revenue generated by generative models with the artists whose data played a role in training these models.

Currently, Runway does not offer an option for opting out of model training or a contributor fund. However, Germanidis suggested that the company is actively exploring the possibility of implementing these features.

Germanidis emphasized the significance of Runway’s creative foundation, underscoring how it has shaped the company’s ethos and its approach to advancing this technology. He expressed the company’s strong commitment to ensuring that artists find value in their products and tools, aligning them with the creative community’s needs and aspirations.

However, there’s another facet to the copyright debate worth exploring: the copyrighting of AI-generated works. The question of whether AI-generated works can be subject to copyright remains unanswered. The U.S. Copyright Office has only recently started seeking input and comments on matters related to generative AI and intellectual property, and judicial rulings have yet to provide substantial clarity on this complex issue.

Germanidis confidently stated that content generated through Runway can indeed be copyrighted. While he didn’t go as far as committing to a policy akin to Microsoft’s recent initiative, which covers legal damages related to copyright for Microsoft AI services users, he assured that Runway is prepared to support its customers if the need arises.

“We’re ready to adjust to any regulatory changes necessary,” he asserted. “Artists should have full confidence in using our platform. We firmly stand by the content they produce, and it unequivocally belongs to them.”

Scenario 1: Copyright Clarity and Collaboration

In this scenario, the world of generative AI witnesses a collaborative effort among governments, AI companies, and artists. They work together to establish a clear and harmonized global copyright framework designed specifically for AI-generated content. International copyright laws are enacted, addressing the rights and responsibilities of all parties involved. This reduces legal disputes, and AI platforms adopt standardized revenue-sharing models. Artists whose data contributes to AI models receive a fair share of generated revenue, fostering a balanced relationship. Robust mechanisms for artists to opt out of model training are introduced, ensuring creative control. Such collaborative efforts result in a unified global copyright framework, providing clarity and confidence for artists and AI creators.

Outcome: Legal clarity and collaboration lead to a thriving generative AI industry. Artists and creators embrace AI platforms, confident in their rights and creative control. Innovation flourishes as legal barriers diminish, enabling cross-border collaboration and a vibrant AI ecosystem.

Scenario 2: Ongoing Copyright Ambiguity

In this alternate scenario, the copyright landscape for generative AI remains uncertain. Legal battles continue as artists sue AI companies for alleged copyright infringements. Ownership of AI-generated content remains unclear, causing disputes and mistrust. Some AI platforms lack effective opt-out mechanisms, leading to artist dissatisfaction. Global consensus on AI copyright standards is absent, resulting in regulatory fragmentation and discouraging investment. Legal ambiguity hampers innovation, leading to stagnation in the generative AI domain.

Outcome: Legal uncertainty hinders generative AI growth. Artists are cautious about AI platforms, and innovation stalls due to legal challenges. The absence of a unified copyright framework impedes progress in the generative AI industry.

Conclusion

In conclusion, the copyright issues surrounding generative AI are complex and far from being resolved. Companies like Runway are working closely with artists to find the best approaches to address these concerns and ensure that their products and tools serve the creative community’s needs. While Runway currently does not offer mechanisms for artists to opt out of model training or a contributor fund, the company is actively exploring the possibility of implementing these features.

Moreover, Runway asserts that content generated through its platform can be copyrighted, and the company is ready to support its customers in defending their rights if necessary. As the debate around copyrighting AI-generated works continues, Runway remains committed to empowering artists and ensuring that their content unequivocally belongs to them.

Do you think AI-generated works should be copyrighted? How do you see generative AI impacting the world of creativity and copyright in the coming years? Share your insights below.