Major Studios Oppose New AI Copyright Legislation, Assert Existing Law is Sufficient
In the ongoing discussions surrounding artificial intelligence (AI) and copyright, major studios caution against implementing rigid rules. According to a filing by the Motion Picture Association (MPA) with the U.S. Copyright Office, the studios believe that current copyright laws are adequate to address the challenges posed by AI technology.
The MPA’s legal team, including Karyn Temple, Benjamin Sheffner, and Terrica Carrington, stated in the filing that existing copyright doctrines and principles should be sufficient to handle any AI-related copyright questions that may arise. They further emphasized that, at this time, there is no need for new legislation or special rules specifically for AI.
The studios maintain that AI has the potential to enhance human creativity rather than supplant it. They believe AI can coexist with a copyright system that encourages original expression and safeguards copyright owners’ rights.
While President Joe Biden recently issued an executive order addressing AI concerns such as deepfakes, copyright matters will need to be resolved by Congress. Senate Majority Leader Chuck Schumer (D-NY) has been involved in crafting AI legislation, and draft bills for unauthorized use of a performer’s digital likeness have been proposed.
Within the filing, the MPA expressed disagreement with the Copyright Office’s stance on protecting content that contains AI-generated material. The Copyright Office stated that only material humans create is eligible for copyright protection. The MPA argued that the Office has not adequately distinguished between generative AI and routine post-production AI tools.
The MPA also addressed using copyrighted material in AI training models, citing cases where authors like John Grisham and George R.R. Martin have sued OpenAI and Meta for incorporating their works in training AI systems. The MPA suggested that potential infringement cases should be evaluated based on the fair use doctrine, considering factors such as the purpose of the use, the nature of the work, the amount used, and the impact on the market.
Overall, the MPA urged for a comprehensive examination of AI’s impact on copyright, cautioning against rigid rules that fail to recognize the nuances and complexities of human creativity intertwined with AI tools.
The Motion Picture Association (MPA) is taking a stance on using copyrighted works in AI training models. They argue that opt-in consent should be required to use owners’ works for training, emphasizing the importance of protecting copyright owners’ exclusive rights.
While some AI companies like OpenAI have licensing agreements with Shutterstock, the MPA warns that alternative methods, like opt-out systems, may not be effective. The sheer scale and volume of works in MPA members’ libraries make opt-out regimes burdensome and insufficient. Additionally, these solutions fail to address the problem of pirated content used as training material.
However, the MPA does see value in developers keeping records of the materials used in their models. This record-keeping could be crucial in potential litigation cases.
The MPA’s approach to AI reflects their ongoing concerns over copyright protection and piracy in the digital era. They understand the need for flexibility in the industry and acknowledge the studios’ interest in creator-driven tools, including AI technologies.
On the topic of proposed legislation, such as the No Fakes Act and protections for artistic style, the MPA expresses reservations. They believe that stringent regulations on artistic style may hinder creative freedom and lead to vague and burdensome litigation.
Furthermore, President Biden’s executive order aims to establish standards for authenticating content and watermarking AI-generated materials. Studios oppose mandatory labeling or disclosure requirements for works that utilize AI-generated material for expressive and entertainment purposes, as they believe it would impede creative freedom. However, they support labeling and identification in cases where AI is intentionally designed to deceive or mislead for political or other reasons.
In summary, the MPA advocates for consent and proper documentation in using copyrighted works for AI training models while considering the studios’ desire for creative freedom in developing AI technologies. They urge caution in implementing regulations that may limit artistic expression or impose unnecessary labeling requirements.