前 Stability AI 音频副总裁 Ed Newton-Rex 最近推出了一项名为 Fairly Trained 的计划，旨在为那些在训练过程中未侵犯版权的人工智能模型授予认证。这一创新举措旨在推动AI行业的透明度和合法性，确保公司在使用受版权保护的数据时遵守法律。
Fairly Trained 计划的首个认证被称为“许可模型认证”（Licensed Model certification），专门针对那些已获得使用受版权保护的训练数据许可的公司。据《The Verge》报道，目前已有9家专注于图像、音乐和语音生成的生成式人工智能公司成功获得了这一认证。
这一认证的推出，标志着AI行业在版权保护方面迈出了重要一步。通过此认证，消费者和企业可以更加信任这些AI模型，知道它们在生成内容时遵循了合法的版权规定。此举不仅有助于保护原创内容创作者的权益，也有助于构建一个更加健康、公正的AI发展环境。
Ed Newton-Rex 表示，Fairly Trained 计划的目的是消除AI模型训练过程中的法律不确定性，鼓励业界采取负责任的实践。随着AI技术的广泛应用，确保模型训练的合法性和透明度变得日益重要。
获得“许可模型认证”的公司可以展示其在AI创新方面的合规性，同时也向市场传达了其对版权尊重的承诺。这一计划有望成为行业标准，推动其他AI企业跟进，共同构建一个尊重知识产权的未来。
英语如下：
**News Title:** “Fairly Trained Initiative Launches: 9 AI Companies Receive First ‘Licensed Model Certification'”
**Keywords:** AI certification, licensed model, copyright protection
**News Content:**
Title: Fairly Trained Initiative Takes Off, Awarding Certification to AI Models Respecting Copyrights
Former Stability AI Audio Vice President Ed Newton-Rex has recently unveiled the Fairly Trained program, aimed at certifying artificial intelligence (AI) models that have not infringed on copyrights during their training process. This innovative move seeks to enhance transparency and legitimacy in the AI industry, ensuring companies abide by the law when using copyrighted data.
The first certification under the Fairly Trained program is the “Licensed Model certification,” specifically targeting companies that have obtained permission to use copyrighted training data. According to _The Verge_, nine companies specializing in image, music, and voice generation AI have successfully secured this certification.
The introduction of this certification marks a significant stride in the AI industry’s efforts towards copyright protection. Consumers and businesses can now have greater trust in these AI models, knowing that they adhere to lawful copyright regulations when generating content. This not only safeguards the rights of original content creators but also contributes to fostering a healthier and more equitable AI landscape.
Ed Newton-Rex states that the Fairly Trained initiative aims to eliminate legal uncertainties in AI model training and promote responsible practices. As AI technology becomes increasingly widespread, ensuring the legality and transparency of model training is becoming more crucial.
Companies receiving the “Licensed Model Certification” demonstrate their compliance with AI innovation regulations and signal their commitment to respecting copyrights. This initiative is poised to set an industry standard, encouraging other AI firms to follow suit and collectively build a future that values intellectual property.
【来源】https://www.theverge.com/2024/1/17/24041518/generative-ai-copyright-violation-fair-training-label-certification
Views: 1