Input out, output in: towards positive-sum solutions to AI-copyright tensions
Abstract This article addresses the legal tensions between artificial intelligence (AI) development and copyright law, exploring policymaking on the use of copyrighted data for AI training at the input level and the generation of AI content at the output level. Currently, global policy responses have heavily focused on the input level—whether AI can lawfully incorporate copyrighted data for AI training purposes. Jurisdictions such as the EU, the UK, the US, China and Japan adopt varied approaches. By comparing and examining different policy versions, this article proposes shifting the focus from input restrictions to output regulation, a policy strategy referred to as ‘input out, output in’. It suggests that AI training should generally be lawful, while regulatory guardrails should apply to outputs that may compete directly with copyrighted works and deprive rightsholders of their deserved revenues. To harmonize the relationship between copyright holders and AI developer
Abstract This article addresses the legal tensions between artificial intelligence (AI) development and copyright law, exploring policymaking on the use of copyrighted data for AI training at the input level and the generation of AI content at the output level. Currently, global policy responses have heavily focused on the input level—whether AI can lawfully incorporate copyrighted data for AI training purposes. Jurisdictions such as the EU, the UK, the US, China and Japan adopt varied approaches. By comparing and examining different policy versions, this article proposes shifting the focus from input restrictions to output regulation, a policy strategy referred to as ‘input out, output in’. It suggests that AI training should generally be lawful, while regulatory guardrails should apply to outputs that may compete directly with copyrighted works and deprive rightsholders of their deserved revenues. To harmonize the relationship between copyright holders and AI developers, key policy tools may include promoting transformative use, proper quotation and attribution, a Creative Commons-style framework and the safe harbour mechanism. This output-focused approach seeks to create positive-sum outcomes for copyright holders, AI developers and public information consumers. By ensuring free access to training data while moderating AI-generated content, the proposal supports innovation, protects creators’ interests and enhances public access to quality information, ultimately promoting a balanced and sustainable legal framework.
Executive Summary
This article proposes a novel approach to resolving AI-copyright tensions by shifting focus from input restrictions to output regulation. The 'input out, output in' strategy suggests that AI training should be generally lawful, while regulatory guardrails apply to outputs competing with copyrighted works. This approach aims to create positive-sum outcomes for copyright holders, AI developers, and public information consumers by promoting innovation, protecting creators' interests, and enhancing public access to quality information.
Key Points
- ▸ The current global policy responses focus on the input level, with varying approaches across jurisdictions
- ▸ The proposed 'input out, output in' strategy prioritizes output regulation over input restrictions
- ▸ Key policy tools include promoting transformative use, proper quotation and attribution, a Creative Commons-style framework, and the safe harbour mechanism
Merits
Encourages Innovation
The proposed approach supports innovation by allowing free access to training data, which is essential for AI development
Demerits
Complexity in Implementation
The output-focused approach may be challenging to implement, as it requires nuanced regulations to balance the interests of copyright holders and AI developers
Expert Commentary
The article's proposal offers a thoughtful and innovative solution to the AI-copyright conundrum. By prioritizing output regulation, the approach acknowledges the importance of protecting creators' interests while promoting innovation and public access to information. However, the implementation of such a strategy will require careful consideration of the complex interplay between copyright law, data protection, and AI development. As the legal landscape continues to evolve, it is essential to engage in ongoing dialogue between stakeholders to ensure that the regulatory framework supports the needs of all parties involved.
Recommendations
- ✓ Policymakers should consider adopting a flexible and adaptive regulatory approach to address the rapidly changing AI-copyright landscape
- ✓ Copyright holders and AI developers should engage in collaborative efforts to develop and implement the proposed 'input out, output in' strategy
