Chinanews.com, Beijing, February 9th (Reporter Yuan Xiuyue) Recently, the artificial intelligence chat robot ChatGPT under the American OpenAI company has become a phenomenon-level application that is popular all over the world.
As an emerging digital technology, artificial intelligence is influencing the world around us with its powerful force.
When people are curious, excited, or worried, resisting and other emotions, how to deal with legal issues brought about by artificial intelligence, and to be alert to related legal risks is also particularly important.
webpage Screenshot
Who does the content generated by ChatGPT belong to? can it have authorship
At present, people use ChatGPT in a wide range of scenarios, including writing emails, video scripts, copywriting, codes, poems, papers, etc.
A foreign survey of more than 100 educators and more than 1,000 students showed that more than 89% of students use ChatGPT to help with homework.
Some researchers also use it to write academic papers. For this reason, several academic journals have changed their editing rules and prohibited ChatGPT from being listed as co-authors.
So, legally, can ChatGPT have authorship?
Who owns the copyright of the text content it generates?
Li Weihua, a partner of Dacheng Law Firm, told a reporter from Chinanews.com that the "author" referred to in my country's "Copyright Law" is a natural person, a legal person or an unincorporated organization, and AI is not an "author" in my country's "Copyright Law". According to the law, ChatGPT is directly granted authorship.
Even for the few countries that have explicitly passed legislation to recognize that AI-generated content can obtain copyright protection, they usually grant the developer of computer software, or the person who has made substantial contributions to the generation process, the authorship.
In addition, he mentioned that it also depends on the agreement between the computer software developer and the user.
For example, the copyright division of works generated by ChatGPT can be learned from its terms of service.
Subject to OpenAI's Terms of Use, OpenAI assigns all of its right, title, and interest in Output Content to the user providing Input Content provided that User complies with the Terms of Use, which include certain restrictions on use.
However, due to the characteristics of machine learning, different users may obtain the same or similar output content from AI, and according to the terms, only the exclusive "output content" rights can be transferred.
Therefore, the transfer of this output content needs to exclude the results presented to other ChatGPT users, and such restrictions will cause uncertainty about the copyright of the output content, and it will become more unclear as time goes by.
Data map: Beijing, in the Shougang Park exhibition area of the Service Trade Fair, dancing robots in the telecommunications, computer and information service exhibition hall.
Photo by Chinanews.com reporter Li Jun
Can works generated by artificial intelligence be regarded as original works?
The use of artificial intelligence for content creation is not new, but with the development of technology, the production level of artificial intelligence content has been continuously improved, and there are even cases where AI paintings have won awards.
Can works generated by artificial intelligence be regarded as original works?
Li Weihua said that Article 3 of my country's "Copyright Law" stipulates that the works referred to in this law refer to intellectual achievements that are original and can be expressed in a certain form in the fields of literature, art and science.
Therefore, in theory, no matter the content created by a natural person or the content generated by artificial intelligence such as ChatGPT, as long as it meets the above-mentioned constituent elements, it can constitute a work.
However, at present, Chinese laws do not clearly stipulate whether the content generated by artificial intelligence can constitute a work, and Chinese courts also have different understandings of this in judicial practice.
In this regard, Li Weihua believes that as long as the content generated by artificial intelligence meets the constitutional requirements of works in the Copyright Law, it should be granted copyright protection.
"Because these contents that seem to be created by computers or artificial intelligence are fundamentally the result of human intelligence. Computers or artificial intelligence are only calculating and outputting according to the programs, content or algorithms set by humans in advance." He said .
What are the legal risks?
The development of artificial intelligence has also brought a series of new legal issues. For ordinary users, what legal risks should be paid attention to?
In Li Weihua's view, the first is the risk of intellectual property infringement, because ChatGPT is a large-scale language model trained on a large number of different data sets, and using copyrighted materials to train the artificial intelligence model may cause the model to provide responses to users. Sometimes excessive reference to other people's works, resulting in infringement disputes.
Li Weihua reminded that users should be aware of this risk, and consider feasible measures such as avoiding narrowing chat requests to avoid or limit access to output content that tends to be more precise and narrowly defined that may come from specific works.
The second is data compliance risk.
OpenAI's terms of use stipulate that OpenAI has broad rights to use any user input and output content to improve ChatGPT.
On this basis, if the user enters various information related to identified or identifiable natural persons (not anonymized) when entering a request, there is a risk of infringing on the rights and interests of other people's personal information.
Therefore, when using ChatGPT, users should anonymize all data that may involve personal information, or avoid typing any content that involves personal information.
Finally, there is the risk of disclosure of trade secrets and breach of confidentiality obligations.
Li Weihua explained that users may enter confidential information when submitting a request without precaution or vigilance.
OpenAI's terms of use do not provide any protection for confidential information that users may enter into ChatGPT.
Instead, the terms of use state that OpenAI has broad use rights to any user input and output content to improve ChatGPT, but allow users to contact OpenAI to choose not to allow OpenAI to use the content in question for such purposes.
Therefore, when using ChatGPT and similar artificial intelligence tools, users, especially those with confidentiality obligations, should pay attention to taking measures to avoid leaks.
(over)