ChatGPT (GPT standing for ‘Generative Pre-Trained Transformer’) is a large conversational dialogue and language model developed by OpenAI (an artificial intelligence research non-profit company, established in 2015 by Twitter owner Elon Musk, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, and Sam Altman)which makes research and access to information smooth-sailing. The AI-powered chatbot came into the public domain on 30th November, 2022 and the free GPT-3.5 language technology has generally been hailedthe most important and useful advance in AI for years. You only need to bang out your query in its search bar and- boom- the detailed answers are populated within a fleeting moment. You are even amazed to discover that even the most complicated questions are answered in a way that only persons with expert knowledge would answer. Generally, the results you get when you use the ‘machine’ are leaps and bounds ahead of what has ever been seen in AI. This fact prompted us to ask ChatGPT whether its style is to ‘copy and paste’ materials from the internet, an attempt which is likely to infringe on certain Intellectual Property rights, and we got the following response:
ChatGPT does not simply copy content from the internet. It is a large language model that has been trained using a vast amount of text data, including books, articles, and other written material. This allows ChatGPT to generate responses based on the context of a conversation in a way that sounds natural and human-like. While ChatGPT may use information from the internet to generate its responses, it does not simply copy and paste content from websites.”
However, as the use of ChatGPT and other artificial intelligence becomes more widespread, it is important to consider the impact it may have on laws related to copyright, trademark, and intellectual property in Nigeria. Well, we sought to know further how ChatGPT works, first off, and the best way to have a clear-cut answer to the question is to turn to ChatGPT itself. Listen:
“ChatGPT is a chatbot that uses a large language model trained by OpenAI. It is a variant of GPT-3 (Generative Pretrained Transformer 3), which is a state-of-the-art natural language processing model. ChatGPT has been specifically trained to respond to human input in a conversational way, and it can generate human-like text based on the context of the conversation. This allows ChatGPT to engage in natural-sounding conversations with users on a variety of topics. ChatGPT can be used in a variety of applications, including customer service, chatbots, and virtual assistants.“
It is generally known that copyrighted materials are a creation of human authors, and such “original works”, fixed in tangible form, must be expressed in a physical or digital form, such as a book or painting so as to be covered by copyright law. But how can we say the same of the content generated by ChatGPT? Can the content be said to be original and fixed in a tangible form, and therefore eligible for copyright protection? ChatGPT weighs in:
“ChatGPT is a computer program created by OpenAI, and as such, it is considered a type of literary work under copyright law. The code and the output generated by the program may be eligible for copyright protection, as it is fixed in a tangible form, such as on a computer’s hard drive or in a cloud-based storage system. However, the specific details of copyright protection for ChatGPT, as well as other AI-generated works, are complex and may be subject to debate and interpretation by legal experts.”
Another question posed is how legal it is to copy materials from ChatGPT for use in commercial purposes. ChatGPT breathlessly proclaims:
“It is generally legal to use the content produced by ChatGPT in commercial use cases, provided that you have the appropriate permissions and licenses. As a large language model trained by OpenAI, ChatGPT generates text based on the context of a conversation, and the specific responses it produces will vary depending on the input it receives. If you want to use the content generated by ChatGPT for commercial purposes, you may need to obtain a license from OpenAI or the relevant rights holders, depending on the specific circumstances of your use case. It is always important to ensure that you have the necessary permissions and licenses before using any content for commercial purposes.”
It is without doubt that the use of ChatGPT has been endorsed by great demonstrations of popular enthusiasm, and as more people begin to use it for different purposes, the risk looms so large that ChatGPT may unwittingly generate content that infringes on the IP rights of others or repetitive of other AI-generated content, which brings the question to the fore as to whether ChatGPT can be sued for breach of Intellectual Property in Nigeria. Listen to ChatGPT’s tightly disciplined response:
“It is possible for ChatGPT or its parent company, OpenAI, to be sued for intellectual property breaches in Nigeria if it is found to have used or distributed copyrighted or trademarked materials without permission. However, whether or not a lawsuit is successful would depend on the specific circumstances of the case and the laws and regulations in place in Nigeria at the time. It’s worth noting that as a machine learning model, ChatGPT is not capable of intentionally committing IP breaches, and any such breaches would likely be caused by the actions of its users or operators.”
A corollary to that question is whether ChatGPT can be sued at all?
“As a machine learning model, ChatGPT does not have the legal capacity to be sued. It is a creation of OpenAI and any legal action would be directed towards the company. OpenAI can be held liable for any damages caused by the model, if it can be shown that the company was negligent in its development or deployment. However, proving negligence in the case of AI can be difficult.
It’s also worth noting that OpenAI has implemented safety measures to ensure that the model is used responsibly, and provides guidelines and terms of use for the model to protect users, and limit the liability of the company.”
When asked how ChatGPT can protect intellectual Property Rights in Nigeria, it merely delivers a heavy homily replete with blank advice:
“As a machine learning model, ChatGPT does not have the ability to actively protect intellectual property rights. However, the company that created and maintains the model, OpenAI, can take steps to ensure that the model is used in a way that respects intellectual property rights. Some ways OpenAI may do this include:
Providing clear guidelines and terms of use for the model that prohibit the use of copyrighted or trademarked materials without permission.
Regularly monitoring the model’s output to detect and remove any instances of infringement.
Providing tools for users to report instances of intellectual property infringement.
Educating users about intellectual property laws and regulations in Nigeria and around the world.
Taking legal action against any individuals or organizations that use the model to infringe on intellectual property rights.
It’s also worth noting that OpenAI may also have a legal obligation to take steps to prevent infringement under Nigerian laws, including the Nigerian Copyright Act, Trademarks Act and Industrial Designs Act.”
When we also asked whether there are possible legal issues that might arise with the use of ChatGPT in intellectual property laws, the A1 model cautions:
“There are several potential legal issues that may arise with the use of ChatGPT in relation to intellectual property laws, including:
Copyright infringement: ChatGPT generates text based on its training data, which may include copyrighted materials. If the model is used to generate text that infringes on the copyright of another, the user or company operating the model may be liable for copyright infringement.
Trademark infringement: ChatGPT may generate text that includes trademarks owned by others. If the model is used in a way that creates confusion or dilutes the distinctiveness of a trademark, the user or company operating the model may be liable for trademark infringement.
Trade secret misappropriation: ChatGPT’s training data may include trade secret information belonging to third parties. If the model is used to generate text that reveals trade secret information, the user or company operating the model may be liable for trade secret misappropriation.
Right of publicity: ChatGPT may generate text that includes the names or likenesses of real people. If the model is used in a way that invades the privacy or misappropriates the identity of a person, the user or company operating the model may be liable for violating the right of publicity.
It’s worth noting that these legal issues arise when the model is used in a way that infringes on the intellectual property rights of others, and the responsibility lies with the user or the company operating the model, and not the model itself.”
Some tweeps argue ably, though tediously, that ChatGPT would soon render legal advice from human lawyers obsolete or unnecessary, ChatGPT dismisses this notion, saying:
“As a machine learning model, I can provide helpful information and assist with answering questions, but I am not a substitute for professional legal advice from a qualified attorney. My knowledge is based on the text data that I’ve been trained on, and my understanding of the law is limited by that data. Additionally, the law is complex and can vary depending on the jurisdiction, therefore, the interpretation of the law and the application of it to specific situations is best done by a human lawyer who is familiar with the applicable laws and regulations.
It’s also worth noting that while I can provide helpful information and assist with answering questions, I can’t replace the human judgement, discretion, and creativity that a human lawyer bring to legal matters. Furthermore, laws and regulations are constantly changing and there may be nuances and subtleties that I am not aware of.
In summary, my goal is to assist and provide helpful information, but I should not be considered a substitute for professional legal advice from a qualified attorney.”
It also bears asking why ChatGPT does not include citations or attributions to the original sources and Intellectual Property used. It says:
“ChatGPT is a machine learning model that generates text based on patterns it has learned from a large dataset of text. It is not designed to include citations or attributions in its responses. This is because the model does not have the capability to determine the origin of the information it is providing or to assess the accuracy or reliability of that information. Additionally, citations and attributions are typically added by humans who are manually curating content and assessing its credibility.”
And since ChatGPT does not want to be cited as a source, say “ChatGPT-generated response, accessed 17/1/2023”, it adds:
“ChatGPT is a machine learning model and not a source of information in the traditional sense, so it would not be appropriate to cite it as a source in a research paper or other document. If you use information generated by the model in your work, you would need to provide context for how the information was generated (e.g., “generated by ChatGPT, a language model trained by OpenAI”) and include a disclaimer that the information is not independently verified. Additionally, it is important to keep in mind that information generated by a language model like ChatGPT may not be accurate or reliable and should be fact-checked before use.”
Finally, it has been said, not truly, but with a possible approximation to truth, that even ChatGPT also sometimes makes odiously embarrassing mistakes, and that some of its responses are sometimes inaccurate. ChatGPT itself admits that we should not always be obedient to its controlling intelligence, and that any notion of infallibility on its part is an optical illusion of some sort, adding that it is not everything it says we can gobble with unabashed glee:
“As a machine learning model, ChatGPT can make mistakes. These mistakes can occur for a variety of reasons, including:
Limited training data: ChatGPT is trained on a large dataset of text, but it’s impossible to cover every possible scenario and context. If the model encounters an example that is not represented in its training data, it may make an error in its output.
Data bias: The data that ChatGPT is trained on may include biases that are reflected in the model’s output. For example, if the training data contains more examples of a certain group of people, the model may be more likely to make errors when generating text about other groups.
Ambiguity: The natural language can be ambiguous and ChatGPT may make errors when interpreting the meaning of a sentence or phrase.
Lack of context: In some cases, the model may generate text that is accurate but is not relevant or appropriate for the given context.
Unforeseen use: The model may make errors when used in ways that were not anticipated or for which it wasn’t specifically trained.
As a language model, ChatGPT is continuously updated and improved to minimize errors, but it’s important to keep in mind that it’s not perfect and that any information generated by the model should be verified and fact-checked before being used.”
0 Comments