美国专利商标局发起讨论 人工智能的创作成果能否获得知识产权？

该办公室正在收集有关人工智能对版权，商标和其他知识产权的影响的信息。它概述了13个具体问题，范围从AI创作侵犯版权的作品到合法提供AI版权材料是否合法。

首先要问的是，人工智能在没有人类任何创造性参与的情况下所产生的输出是否应被视为受美国版权法保护的作者作品。如果不是，那么人类参与的程度“是否或应该足以使作品符合版权保护的条件？”

其他问题则询问培训AI的公司是否应该拥有最终的作品，以及是否可以首先使用受版权保护的材料来培训AI。办公室问：“作者是否会因其作品的这种使用方式而受到认可？” “如果是这样，怎么办？”

该办公室除了向政府提供版权方面的建议外，还经常征询公众意见，以了解最新动态并听取与之实际交流的人的意见。今年早些时候，该办公室同样要求公众对AI和专利发表意见。
““如果确实是按钮操作，并且您得到了结果，那么我认为该内容没有任何版权。”

这些问题在美国法律中都没有具体答案，但是多年来人们一直在争论潜在的结果。当您查看基于AI的应用程序之类的东西时，情况可能会更清楚一些，在该应用程序中，用户必须做出很多决定才能确定最终结果。乔治华盛顿大学法学院讲师兹维·罗森（Zvi S. Rosen）告诉《边缘报》：“我认为，可以保护的是，一个人为参与作者身份而采取的有意识的步骤。”但是，如果有人使用一次单击即可显示结果的AI，那就另当别论了。 “我的看法是，如果这确实是一个按钮式的事情，您会得到一个结果，但我认为其中没有任何版权。”

但是它并不总是切成薄片和干燥的。编码人员已经宣称他们的AI软件创建的“按钮”作品是作者的，这是在今年早些时候通过与初创公司Endel达成的华纳音乐发行协议达成的。罗森说：“那就是事情变得更加复杂。” “对此我没有明确的答案。”

正如The Verge先前所检查的那样，此类问题是围绕AI和版权法进行持续讨论的核心。这是一个彻头彻尾的混乱主题，没有明确的答案。 《美国版权局实践纲要》中有一些基本指南，其中指出，如果机器在没有创造性输入或没有人工干预的情况下生产的作品，将无法获得作者权。但是，随着AI在创意作品中的作用越来越复杂和细微差别，专利商标局似乎无法接受这个定义。
“人工智能的创作将继续变得更加复杂和细致”

通常，USPTO在进行此类查询时只会得到公众的回应，大部分来自律师事务所，公司和各种利益集团。但是任何人都可以发表评论，罗森说，这对个人创作者有所帮助。他说：“ [办公室]正在寻找特定的问题和互动，因此，例如，如果音乐家曾与AI合作过并且可以证明特定的经历或不满，那就很有帮助。”

随着AI变得越来越先进，在创意工作流程中司空见惯并能够制作自己的材料，办公室提出的问题已不再是理论上的问题，必须予以回答。罗森说：“（办公室）这样做是不足为奇的。” “我认为每个人都看到它的到来。考虑到这些事情要花多长时间，任何立法回应都将迟到，但是通过在研究结束时力争做到这一点，就不会迟到。事情就是这样。”

问题的完整列表，以及有关如何通过电子邮件将您的意见发送到办公室的说明，请参见《联邦公报》通知。评论期于12月16日关闭。

原文：

The US office responsible for patents and trademarks is trying to figure out how AI might call for changes to copyright law, and it’s asking the public for opinions on the topic. The United States Patent and Trademark Office (USPTO) published a notice in the Federal Register last month saying it’s seeking comments, as spotted by TorrentFreak.

The office is gathering information about the impact of artificial intelligence on copyright, trademark, and other intellectual property rights. It outlines thirteen specific questions, ranging from what happens if an AI creates a copyright-infringing work to if it’s legal to feed an AI copyrighted material.

It starts off by asking if output made by AI without any creative involvement from a human should qualify as a work of authorship that’s protectable by US copyright law. If not, then what degree of human involvement “would or should be sufficient so that the work qualifies for copyright protection?”

Other questions ask if the company that trains an AI should own the resulting work, and if it’s okay to use copyrighted material to train an AI in the first place. “Should authors be recognized for this type of use of their works?” asks the office. “If so, how?”

The office, which, among other things, advises the government on copyright, often seeks public opinion to understand new developments and hear from people who actually deal with them. Earlier this year, the office similarly asked for public opinion on AI and patents.
"“if it’s really a push button thing, and you get a result, I don’t think there’s any copyright in that.”"

None of these questions have concrete answers in US law, but people have been debating the potential outcomes for years. The situation might be a little clearer when you’re looking at something like an AI-based app where a user has to make a lot of decisions to shape the end result. “I think what’s protectable is conscious steps made by a person to be involved in authorship,” Zvi S. Rosen, lecturer at the George Washington University School of Law, tells The Verge. But if someone uses an AI that spits out a result with a single click, that could be a different matter. “My opinion is if it’s really a push button thing, and you get a result, I don’t think there’s any copyright in that.”

But it’s not always cut and dry. Already, coders have claimed authorship over the “push button” works their AI software creates, which happened earlier this year in a distribution deal Warner Music brokered with the startup Endel. “That’s where it gets more complicated,” Rosen says. “I don’t have a clear answer on that.”

As The Verge previously examined, questions like these are at the heart of ongoing discussions around AI and copyright law. It’s a downright messy subject with no clear answers. There is some basic guidance in the Compendium of US Copyright Office Practices, which says that works produced by a machine with no creative input or intervention from a human can’t be given authorship. But it looks like the Patent and Trademark Office feels this definition won’t hold up as AI’s hand in creative works continues to get more complicated and nuanced.
"AI’s hand in creative works continues to get more complicated and nuanced"

Usually, the USPTO only gets a few responses from the public when it makes these types of inquiries, with the bulk coming from law firms, companies, and various interest groups. But anyone can send in a comment, and Rosen says it would be beneficial for individual creators to contribute. “[The office] is looking for specific concerns and interactions,” he says, “so if, for example, a musician has worked with AI and can attest to a particular experience or grievance, that’s helpful.”

As AI becomes increasingly advanced, commonplace in the creative workflow, and capable of making its own material, the questions the office is posing have ceased being theoretical and will have to be answered. “It’s not surprising [the office] is doing this,” says Rosen. “I think everyone sees it coming. Given how long these things take, any legislative response is going to be late, but by trying to get out in front of it on the study end, it’s not going to be as late. That’s just how things work.”

The full list of the questions, along with directions on how to email your comments to the office are available in the Federal Register notice. The comment period closes on December 16th.