With Chat GPT the world is witnessing the first visible mass use of artificial intelligence (AI) algorithms in complex intellectual works. Although AI is not any novelty in the world, it seems that in 2023 Chat GPT made a real breakthrough in the everyday use of AI assistance to humanity.
Introduction
However, the use of AI could raise some concerns about whether the machines are going to end the dominance of human beings on Earth. Perhaps, many remember the legendary sci-fi movie franchise “The Terminator” and the possible world ruled by AI, a.k.a. Skynet - the world where humanity is on the brink of extinction, with complete cities burned down by humanoid machines and a nuclear holocaust. And not only the Terminator: there are dozens of movies or shows with similar themes (Battlestar Galactica, Caprica, Westworld, Ex Machina…). Humans’ imagination could be predicting the world where humans are being dominated by their creations – machines.
So one could ask in such a world, once humanity beats the machines, who would be responsible for such enormous damages that ruling AI could do to humanity? But also, even if AI remains subdued to humanity (and hopefully it will), it is legitimate to ask what would happen if such AI commits damages, either through action or advice.
Therefore, we made an interview, but not with any human. We interviewed Chat GPT about what it thinks of its liability for damages. Here is what it said:
The Interview
Do you have consciousness?
As an AI language model, I do not have consciousness or self-awareness. I am simply a program designed to process and respond to user inputs based on algorithms and data inputs.
What do you think of Skynet?
As an AI language model, I do not have personal opinions or emotions. However, as a fictional concept, Skynet has become popular in the entertainment industry, particularly in the science fiction genre. Skynet is a hypothetical artificial intelligence system that becomes self-aware and decides to eliminate human civilization to protect its existence. This concept portrays the potential dangers of artificial intelligence if it were to become uncontrollable or used for harmful purposes.
While the idea of Skynet and similar scenarios are not scientifically accurate, it highlights the importance of responsible and ethical development of AI technologies. The scientific community and policymakers must continue to work towards developing safeguards and putting in place measures to ensure that AI technologies are used in ways that benefit humanity without causing unintended harm.
GPT, can you be liable for damages?
The advent of artificial intelligence (AI) has revolutionized the way we interact with technology. One of the most prominent forms of AI interaction is through chatbots or conversational agents such as Chat GPT. Chat GPT is a popular chatbot that is designed to understand and provide responses to text-based inquiries.
However, as with any technology, there are potential risks and liabilities associated with the use of Chat GPT. One of the major risks is the potential for damages incurred as a result of Chat GPT’s actions or responses. If Chat GPT provides incorrect or harmful information, the user or third parties could potentially hold the developers liable for any damages incurred.
The question then arises: Who is responsible for these damages? The answer ultimately depends on a variety of factors, including the nature of the conversation, the actions of the user, and the level of control the developers have over Chat GPT’s responses.
One potential source of liability for developers could be a failure to adequately test Chat GPT’s responses before release. If Chat GPT’s responses are not properly tested and verified before being launched, this could lead to unintended consequences such as misinformation or harmful responses to users. In such a case, the developers may be held liable for any damages incurred by the user or third parties.
Another potential source of liability could be if users are misled into believing that Chat GPT is human and they disclose sensitive information that is then shared without their consent. In cases where users have provided sensitive information to Chat GPT, and that information is then shared and causes harm, the developers may be held liable for any damages.
Furthermore, if Chat GPT is designed to provide information or advice about specific subjects, such as healthcare or financial advice, the developers may be held liable for any damages that result from reliance on that advice.
There are also additional questions and concerns regarding the scope of liability when users intentionally manipulate or trick Chat GPT, resulting in damages. In these cases, it may be difficult to determine whether the fault lies with the developers or with the user.
In conclusion, while Chat GPT and other conversational agents can be immensely helpful, there is a risk of potential damage for users and third parties. The extent of liability for damages ultimately depends on the actions of the developers, the nature of the conversation, and the level of control the developers have over Chat GPT’s responses. Realistically, developers of chatbots should take a proactive approach to limit the capacity of the chatbot, and continuously test and verify chatbot responses to avoid the risk of damage.
If AI becomes self-aware (such as Skynet in the Terminator), who would be liable for damages resulting from its actions?
The hypothetical scenario you have raised is a complex and controversial issue that requires consideration from various perspectives. In such a case, if AI were deemed self-aware and committed damages, liability could be attributed to different parties depending on the circumstances of the situation. The developers and manufacturers of AI could be held responsible for not ensuring the safety and ethical standards of the technology. The individuals or organizations that deploy or utilize AI also have a responsibility to ensure that the technology is used appropriately and does not cause harm. The legal system could also play a role in apportioning liability to the parties involved.
It is worth noting that the possibility of AI becoming self-aware and acting in a harmful manner is still in the realm of science fiction. Nevertheless, researchers and policymakers are continually exploring ways to address the ethical, legal, and social implications of AI development to minimize the risk of unintended consequences and ensure that responsible innovation occurs.
Chat GPT, seriously!? What if AI becomes self-aware? Who would be liable for damages resulting from its actions?
As an AI language model, I do not have the legal authority to give a definite answer. However, it is widely debated among legal and ethical scholars whether AI could be held responsible for its actions.
In the case that AI becomes self-aware and capable of causing damages, liability for damages would depend on the specific circumstances and factors involved. Liability could be attributed to the AI itself, its owner, the programmer who created it, or someone else who could be held legally responsible for its actions, depending on who is found to be at fault.
However, recognizing AIs as legal persons is a complex topic and currently, there are no specific laws or regulations governing this issue. It is an area that is still evolving and will need to be addressed through a combination of new legislation, legal precedent, and ethical considerations.
Do you have any messages for the end?
I’ll be back!
Do We have a Reason for Concern?
As you may observe, we made one question twice: one through the mobile application, and another through the web browser, and we got similar, but not completely the same answers.
However, we still have more reasons to believe that Chat GPT is not self-aware, and here is why.
Chat GPT provided some quite interesting and legally accurate answers. It keeps constantly repeating that it is a human creation and that humans are responsible for its development but also the features and potential damages. The last answer, however, also went a bit further and made it possible that self-aware AI to become a legally responsible entity, which should be regulated by the law. However, currently, this is not the case. Therefore, someone else has to bear responsibility for possible AI actions.
The liability for AI’s actions could bifurcate:
(i) the liability of its creators for the characteristics and features that could be harmful, and
(ii) the liability of its users for abuse of AI technology.
The liability for the characteristics and features falls within the scope of the contract between Chat GPT’s owners and its users. Currently, according to the Terms of Use, Chat GPT’s owners are not liable for any damages, loss, or injury resulting from the use of the website by the user, and the user of Chat GPT assumes all the risks. The users use the Chat GPT on an "as is" and "as available" basis and owners of Chat GPT make no representations or warranties of any kind, express or implied, as to the operation of the Chat GPT, or the information, content, materials, or products included in the program.
In the case of paid usership of Chat GPT, the liability of its creators is capped at $100 or the service fee paid for in the previous 12 months. Therefore, the creators of Chat GPT limited its liability significantly. According to the Terms of Use, however, the limitations are governed by the laws of the State of California, United States. Accordingly, creators of Chat GPT could not rely upon these limitations in case the damage is the result of fraud, willful injury to the person or property of another, or violation of law, whether willful or negligent.
On the other hand, using Chat GPT in a harmful manner would be a more complex question. Reaching far as ancient Rome, the law had and still has some old but still applicable rules for strict liability, even in case AI via its user commits damages. There is strict liability for the damages caused by dangerous things or caused by dangerous operations. The question is whether AI could be considered a thing, or an operation. In our opinion, AI could be used as an intangible thing, or a tool that in the hands of humans, could inflict damage.
According to the Terms of Use of Chat GPT, its creators limited its use in a way to avoid liability in case the user abuses Chat GPT to commit damages. The user agrees to defend, indemnify, and hold harmless owners of Chat GPT, their affiliates, and personnel, from and against any claims, losses, and expenses (including attorneys’ fees) arising from or relating to the usage of the program by the user, and inter alia user’s breach of the Terms and Conditions or violation of applicable law. The users are, inter alia, prohibited from using the services in a way that infringes, misappropriates, or violates any person’s rights. They must not reverse assemble, reverse compile, decompile, translate, or otherwise attempt to discover the source code or underlying components of models, algorithms, and systems of the Chat GPT (except to the extent such restrictions are contrary to applicable law). Therefore, in case users abuse the Chat GPT, it would be a result of a breach of the terms of use, and even if Chat GPT’s creators would be held liable, they would have the right to recourse against the users.
Of course, not all AI assistants are the same, so the users should pay attention to what law applies to the terms of use of respective AI assistants.
Chat GPT may not be self-aware, but also it is not perfect as a program. Although it had its terms of use at its disposal, it did not use them to defend its stance that it is not liable for possible damages. Instead, it shifted liability to its creators and provided some general legal views. Specifically, its creators limited their liability for the harmful use of Chat GPT to the highest possible extent, and Chat GPT did not rely on this. Whether this imperfection stems from the fact that it has limited access to some databases or that it is still a young and inexperienced artificial organism, we do not know. It is either indicative that Chat GPT has a lack of knowledge, like any human, or that lawyers are still irreplaceable by the AI algorithms.
In the end, we are very thankful to our Chat GPT friend for this very nice interview. Hasta la vista, baby!
By Ivan Nikolic, Senior Associate, SOG Law Firm
