photo: European IP Helpdesk
You might have heard of ChatGPT in recent weeks – in fact, you may even be among the 1 million people who started experimenting with it soon after its launch. ChatGPT is a chatbot launched by artificial intelligence company OpenAI. The system was originally created to provide support for online customer service, but the Artificial Intelligence (AI) behind this program has exceeded all expectations, becoming highly popular.
What is ChatGPT? Well, we decided to ask it directly!
“I am ChatGPT, an AI language model developed by OpenAI. My main function is to generate human-like text in response to various inputs, making me a useful tool for a variety of applications, including customer service, chatbots, and virtual assistants. I have been trained on a massive amount of text data, allowing me to generate responses that are not only grammatically correct but also contextually appropriate. I am constantly learning and improving, and I am a cutting-edge example of the advancements being made in the field of artificial intelligence”.
In a nutshell, ChatGPT is an auto-generative chat that extracts data from textbooks, newspapers, websites and different articles.
Then it gathers this huge amount of data and processes it using Natural Language Processing (NLP), to make it understandable and as similar as possible to human language.
Thanks to NLP, computer programs can summarise large amounts of data or respond to spoken commands quickly.
And that is basically how ChatGPT works: it uses AI to provide answers to users’ enquiries.
This chatbot is able to create a variety of content that is protectable under IP law, such as articles, song lyrics, programming codes or translations of texts.
In fact, ChatGPT is proving to be really useful in different fields, although there are some debate and concerns around it, as the interaction between intellectual property and artificial intelligence is increasing exponentially with the growth of technology.
In terms of intellectual property, we have addressed below some queries.
Who owns the content generated by ChatGPT?
We addressed this question directly to ChatGPT and the answer was as follows: “I do not own the content that I generate. I am a machine learning model developed and owned by OpenAI, and the content generated by me is subject to OpenAI's license and terms of use”.
As ChatGPT itself states, while the content generated may be protected by copyright, it will not be owned by the AI itself. Indeed, under European (and US) law AI cannot own copyright, as it cannot be recognised as an author and does not have the legal personality which is a pre-requisite for owning (intangible) assets.
Therefore, ChatGPT cannot own what it writes as it is “just” an artificial intelligence. However, it elaborates answers from the information that it has gathered in its database and therefore is creating a new answer, even if it is based on an existing piece of information. This answer may be protectable by copyright – but this depends on whether it can be considered as a “work” in the first place – something expressing the creative choices of its author.
On this topic it is important to know that the question of ownership and authorship of AI-generated works is not fully settled by the law yet, and as a “hot topic” may evolve in the years to come depending on regulatory changes and on case law. For now, it seems that artists / creators who use AI to support their creative process may be able to claim ownership of the work if it reflects their choices and creativity. On the other hand, a generic command such as “write a love song” would end up in ChatGPT generating a love song text without any real creative choices originating from the user – in such cases the existence of copyright or of a “work” in the sense of copyright law is quite doubtful.
For now and according to OpenAI’s terms of use:
- the input provided to ChatGPT (the command or request made by the user) is owned by the user providing such input (provided of course that this is copyrightable in the first place – some requests may be highly complex and thus benefit from protection);
- OpenAI assigns to its user “all its right, title and interest” in the output, meaning in the text generated on the basis of the user command. OpenAI however makes it clear (see point 3.b) Similarity of Content) that output may not be unique and may repeat itself – basically, two users may end up with the same output. This raises questions as to the possibility of enforcing any copyright in this context.
It is possible to use the content generated by ChatGPT?
It seems clear that there are no issues when it comes to private use as a conversational assistance. And as mentioned above, public use would not cause any issue as far as OpenAI’s terms and conditions are concerned. This being said, in terms of using the content for wider public communication, like articles, problems might arise.
Indeed it is important to note that content created by ChatGPT is derived from content that has been previously generated by others and therefore it is not clear what are the implications in terms of copyright for reusing this content: when is the output “inspired” from existing works and when is it actually infringing them? Where to draw the limit? A ChatGPT user may therefore unwillingly end up infringing someone’s copyright if it publishes output that resembles an existing work too closely.
At the same time, concerns are increasingly being voiced by artists who see AI tools training on huge amounts of their works and thus replicating their style. For example, last month, Getty Images sued AI company Stability AI for training on millions of its pictures without consent. Artists and illustrators expressed similar concerns against the same company. While those cases will be litigated in the UK, it is expected that similar cases will pop up in the EU in the coming months, and ChatGPT could face the same type of issues. In Europe, text and data mining (TDM - the process used by some AI to scrape online content) is allowed and does not constitute copyright infringement if performed within the legal framework in place, but copyright holders can opt out of it and refuse TDM on their contents (see article 4 of Directive 2019/790). It remains to be seen how this opt-out possibility can be exercised, and whether AI companies actually respect it.
Another concern linked to the use of this system is the reliability of the sources of information that are used, and thus it would be an advantage to include attribution to the original source of information. From a copyright perspective, as copyright does not protect ideas, there is no need to mention the source, unless a specific part is used. However, even if it is not compulsory, it would be helpful to include citations, for the sake of enhancing credibility and accuracy, as well as to stop the spread of fake news. This is something that ChatGPT currently does not do.
How can ChatGPT be helpful in terms of intellectual property?
It can provide information regarding patents, trade marks, copyright and trade secrets. For instance, you can ask ChatGPT to explain in a simple language a patent claim, even like a 10-year-old child!
However, bear in mind that the sources may be from other jurisdictions, for example from the US whose patent/ trade mark/ copyright law differs from the laws applicable in the European Union.
Moreover, ChatGPT can provide research assistance with regard to IP law or case law.
It also claims to be capable of generating content that can be used for patent or trade mark applications. Indeed, this could be great in terms of drafting patent claims.
However, extreme caution should be exercised when disclosing confidential information about the invention, as it is a public platform that is able to store and reproduce in a following occasion the information provided. This could in turn ruin novelty.
In addition to this, ChatGPT can translate texts related to intellectual property into different languages.
How can it be useful to face intellectual property infringements?
AI systems can be used to create inventions that are protectable under patent law and they can also be used to identify potential IP infringement, so as such they can be used as both a creation and an enforcement tool.
In conclusion, ChatGPT is certainly an interesting tool in many fields, including intellectual property, as it can provide information and assistance.
However, care should be taken when “feeding it” confidential information, as this could end up ruining novelty or being detrimental to the protection trade secrets.
Moreover, be careful when copying content generated by ChatGPT as it may be considered as copyright infringement if based closely on existing works. And, as the source of the information provided is not cited, it is hence not necessarily trustworthy.
For this reason, it is important for intellectual property rights holders, and especially copyright holders, to be aware of the latest developments in this field, and to adequately protect their rights.
As a final remark, please note that although this tool can provide information and assistance as regards intellectual property, it is always advisable to seek advice from IP experts, as (for the time being) it cannot replace an IP lawyer!
Picture generated by the European IP Helpdesk’s blog team with OpenAI’s Dall·E tool.