The internet has been abuzz about ChatGPT, the new AI chatbot from OpenAI. ChatGPT is a language model that interacts in a conversational way. The dialogue format makes it possible for ChatGPT to answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.
ChatGPT was trained using Reinforcement Learning from Human Feedback (RLHF). That is, human AI trainers provided conversations in which they played both sides—the user and an AI assistant. The human trainers ranked ChatGPT model responses by quality and fine-tuned the model using the Proximal Policy Optimization (PPO) reinforcement learning algorithm. The result: A conversational AI chatbot that can mimic human-generated responses in an astonishing variety of use cases.
What legal issues may arise from the new generation of AI chatbots?
AI chatbots like ChatGPT can mimic the work of lawyers, with varying degrees of success. It’s easy to imagine AI chatbots carrying out common legal functions like writing contracts and briefs. Suffolk University Law School Dean Andrew Perlman reported using ChatGPT to write a 14-page law article in under an hour.
Machine learning experts believe a large proportion of legal work may be augmented – or replaced – by AI chatbots. Legal analysis is based on vast databases of statutes, case law and opinions. There’s a high degree of logic and structure. All of these things lend themselves well to AI and machine learning algorithms.
Perlman was able to use ChatGPT to:
- Draft a brief to the United States Supreme Court on why its decision on same-sex marriage should not be overturned
- Explain the concept of personal jurisdiction
- Develop a list of deposition questions for the plaintiff in a routine motor vehicle accident
- Create a contract for the sale of real estate in Massachusetts
The AI chatbot carried out all these tasks with impressive results.
Chatbots may provide a valuable social benefit by expanding access to legal services. Perlman said the technology offers “significant potential to address access to justice questions” in making legal services available to people of limited means.
While AI chatbots can be powerful tools to unleash human productivity, there are concerns about how they may affect the legal profession.
Whether courts will accept ChatGPT-generated legal documents – and what caveats and attribution they will require – remains an open question. But it seems likely that AI chatbots will become an important part of legal practice in the near future.
Accuracy is a Concern
There are other legal concerns related to the use of ChatGPT, say machine learning experts. One concern is the reliability of the results. The technology is not perfect. AI chatbots have a tendency to create “hallucinations” – made-up responses in cases where it doesn’t know the true answer.
An AI chatbot may not have the same level of understanding and judgment as a human lawyer when it comes to interpreting legal principles and precedents. This could lead to problems in situations where a more in-depth legal analysis is required.
The bot’s creators note on the OpenAI website that ChatGPT shouldn’t be relied upon for advice and that it “sometimes writes plausible-sounding but incorrect or nonsensical answers.” If a lawyer did that, there could be malpractice consequences.
Further issues arising from results trained on existing prior literature:
- There’s a risk that AI chatbots could infringe on intellectual property rights. ChatGPT is trained on a vast amount of text data, including books, articles, and other copyrighted works. This means that ChatGPT’s outputs may potentially infringe on the copyrights of those works.
- AI chatbots could generate offensive or defamatory content. Unlike a human, it doesn’t have the same ability to understand the context or implications of the words it generates. Sharing content that is offensive or defamatory could lead to legal action against its users.
- Data protection could be an issue. ChatGPT has the ability to share personal data—from its training datasets—with its users. This functionality means it probably breaches most of the world’s data protection laws.
ChatGPT in Action
Finally, lets’ see what ChatGPT has to say about this topic. I asked Chat GPT, “What are the legal issues of using chatbots?” Here is its response:
“There are several legal issues that can arise when using chatbots, depending on the specific application and jurisdiction. Here are a few examples:
Data privacy: Chatbots often collect and process personal data from users, which raises issues around data protection and privacy. In some jurisdictions, such as the European Union, this may be governed by laws like the General Data Protection Regulation (GDPR).
Consumer protection: If a chatbot is used to sell goods or services, it may be subject to consumer protection laws, which require companies to be transparent about the terms of the sale and the nature of the goods or services being offered.
Intellectual property: Chatbots may use software, content, and other intellectual property, which may be protected by copyright, trademark, or patent laws. Companies using chatbots should be mindful of these legal issues and ensure that they have the necessary licenses or permissions to use the relevant intellectual property.
Employment law: If chatbots are used to perform tasks that are normally done by human employees, there may be issues around employment law, such as worker classification and the payment of wages.
Discrimination: Chatbots may be programmed to discriminate against certain groups of people, which could be illegal in some jurisdictions. For example, a chatbot that refuses to provide information or assistance to users based on their race or gender could be in violation of anti-discrimination laws.
It’s important for companies using chatbots to be aware of these legal issues and ensure that their chatbot applications are compliant with relevant laws and regulations.”
Conclusions
ChatGPT is an exciting technology, but its proper use requires a careful understanding of the potential legal risks. As these AI chatbots evolve and gain more widespread adoption, users will need to navigate these legal risks to ensure they are using the technology lawfully and responsibly. Machine learning experts can help understand the capabilities and limitations of the technology. At Sidespin Group, we provide technical expertise to interpret complex technologies.
No tags for this post.