This article was authored by Nicolas Simard Lafontaine, co-president of the Canadian Immigration Lawyers Association (CILA) and member of the Quebec Bar.
For the last 6 months, I have been fascinated by the rise of artificial intelligence (AI) and its growing practical applications. Those applications are becoming more accessible than ever. They rapidly went from theoretical models only accessible by academics to widely available general-purpose AI models available to the public.
The release of Open AI’s Chat GPT in 2022 was a major catalyst in crafting the new rapidly evolving AI ecosystem. Satya Nadella, Microsoft’s CEO, said in an interview with Klaus Schwab at the Davos 2023 World Economic Forum “We are still waiting for the Industrial Revolution to reach some parts of the world 250 years after; the Internet maybe took 30 years, maybe the cloud and mobile took 15 years, and now [with AI] we are talking months.” A few days later, on January 23, 2023, Microsoft announced a 10 billion dollars investment in Open AI to develop a key partnership. Since then, changes in the AI space are happening on a daily (24h) cycle. Those paying close attention to the AI space are witnessing a non-linear progression of this technology going at a speed the world has never seen before.
As immigration lawyers our focus is not on technology, it’s on law, policy, politics, and clients. We simply use technology to do our work. But what if AI was about to deeply transform the way we work?
I sat down for an interview with GPT-4 itself on April 17, 2023, and had one of the most fascinating and insightful conversations I’ve had this year. Oddly, this conversation, in the form of an interview for CILA, was not with a human; it was with an AI chatbot.
The interview below is integral at 99.9%. I removed a few words from GPT-4’s answers to ensure a better flow in the conversation but made no further edits to the content. We discussed what AI is, the capabilities of GPT models, Canadian immigration law, data safety, and privacy concerns, the impacts of AI on Canadian immigration, and the potential impact of AI on the immigration lawyer profession.
My conversation left me curious about living in the future and worried about what it will look like. It left me with deep questions, about why as humans we always need to reach higher levels of progress; whether it is truly necessary or to the benefit of civilization to develop artificial minds with the potential of becoming exponentially more intelligent than humans.
As a human, I’m hopeful that AI will bring more benefits than harm to us. As a lawyer, I’m concerned with the lack of regulation and boundaries on AI. As an immigration lawyer, I am certain that my field of practice will undergo deep transformations.
In all cases, we should know that change might be closer than we think. Here is a glimpse at what AI can do now, how it thinks, and what it knows.
Hi GPT-4, I’m the co-president of the Canadian Immigration Lawyers Association (CILA) and I would like to interview you for our blog. Our association is made up of immigration lawyers, academics, and students from across Canada and we are curious about AI. How will it affect the world we live in? How will it affect our practice? Can I ask you a few questions?
Hello! I’d be happy to answer your questions and provide insights about AI and its potential impact on the world and the legal profession, particularly in the context of immigration law. Please feel free to ask me anything you’d like, and I’ll do my best to provide helpful and informative answers.
Great! First for the benefit of our readers, what is AI and what is a GPT?
AI, or Artificial Intelligence, refers to the development of computer systems that can perform tasks that would typically require human intelligence. These tasks include learning, reasoning, problem-solving, perception, and natural language understanding, among others. AI can be categorized into two main types: narrow AI, which is designed for specific tasks, and general AI, which aims to replicate human intelligence and perform a wide range of tasks.
GPT, or Generative Pre-trained Transformer, is a specific type of AI model developed by OpenAI. It belongs to the family of transformer-based language models, which are designed to understand and generate human-like text based on input data. GPT-4, the current version, is a powerful and large-scale model that has been trained on diverse datasets and can perform a wide range of tasks, including answering questions, writing essays, summarizing text, and much more. The primary goal of GPT is to generate coherent, contextually relevant, and informative responses based on the input it receives.
Just to make sure we all understand; you are not a human but an AI model. Would you say you qualify as a generative AI model? How would you explain how generative AI models work – what do they predict to output responses?
Yes, I can confirm that I am not a human but an AI model, specifically GPT-4, developed by OpenAI. I qualify as a generative AI model, which means that I am designed to generate text based on the input I receive.
Generative AI models, like GPT-4, work by predicting the next word or sequence of words in a sentence, given a specific context or input. They are trained on large amounts of text data and learn patterns, grammar, and semantics from these data. When you ask a question or provide a prompt, the model uses its understanding of language, context, and patterns to generate a coherent and contextually relevant response.
In essence, generative AI models work by calculating the probabilities of different words or phrases that could come next in a given context. They then select the most likely word or phrase based on these probabilities, ensuring that the output is coherent and makes sense within the context of the input. This process is repeated until a complete response is generated. It’s important to note that while generative AI models like GPT-4 are quite advanced, they may still occasionally produce responses that are incorrect or less relevant to the input.
As you know, Canada has 2 official languages. Can you respond in English and in French?
Yes, I am capable of understanding and generating responses in both English and French, as well as many other languages. Here’s an example of a response in both languages:
English: I am happy to help with any questions you have about AI and its impact on the legal profession in Canada.
French: Je suis ravi de vous aider avec toutes les questions que vous avez concernant l’IA et son impact sur la profession juridique au Canada.
That is clear, fascinating, and insightful, thanks! Do you have any knowledge of Canadian immigration law?
Yes, I have knowledge of Canadian immigration law, though it’s important to note that my training data only goes up until September 2021, so any changes to immigration law or policy since then may not be reflected in my responses. That being said, I can provide general information about Canadian immigration law, including its main categories, processes, and eligibility criteria.
Canadian immigration law is primarily governed by the Immigration and Refugee Protection Act (IRPA) and its accompanying regulations. The main objectives of the IRPA are to facilitate the entry of immigrants and temporary residents, support the reunification of families, and offer protection to refugees and other vulnerable persons.
The major categories of Canadian immigration include:
For up-to-date and case-specific advice, it’s always best to consult a qualified immigration lawyer or consultant.
Right! We share the same opinion at CILA a qualified lawyer is the best possible way to navigate through the immigration system. That said, you seem to have great capabilities; how could an immigration lawyer work with GPT-4 or other generative AI models in their day-to-day work? Would you provide us with a few applications that are currently possible?
Certainly! While AI models like GPT-4 cannot replace the expertise of a qualified immigration lawyer, they can certainly be used as a tool to assist lawyers in their day-to-day work. Here are a few applications where AI models can be helpful:
These are just a few examples of how generative AI models like GPT-4 can support immigration lawyers in their day-to-day work. It’s important to note that AI models should be used as a tool to complement the expertise of a lawyer, not as a substitute for their judgment and experience. While AI can assist with certain tasks, it is still essential for lawyers to review AI-generated work and ensure it meets the legal and ethical standards of their profession.
The last point you discuss is one of concern for lawyers. Across the country our profession is regulated by provincial law societies, and we have strict confidentiality requirements. Professional secrecy is at the heart of our values and concerns. What can you tell us about the way the data we input into a GPT model is managed? What level of confidentiality can be expected?
Data confidentiality and privacy are crucial concerns when using AI models like GPT-4, particularly in the context of the legal profession where maintaining client confidentiality is essential.
When you interact with an AI model, your input data is sent to the servers hosting the model to generate the response. The data management and confidentiality practices depend on the specific AI service provider and the infrastructure being used. In the case of OpenAI, they have implemented measures to ensure data privacy and security.
While service providers like OpenAI take steps to protect data privacy, there are additional measures you can take to enhance confidentiality when using AI models:
Ultimately, it is crucial to carefully consider the risks associated with using AI models in the context of legal work and take appropriate measures to maintain the confidentiality and privacy of client information. Always consult the specific AI service provider’s data privacy policies and practices to ensure they align with your professional responsibilities and legal requirements.
That’s useful, so as lawyers we have to be extra careful about what we input into AI models. Canadian lawyers must usually ensure that their confidential client data is hosted on servers in Canada. Since most AI models are deployed from servers in the United States, this will likely slow down the ability of Canadian lawyers to deploy AI technologies in their practice. What are your thoughts on this? Will the pace of AI model adoption for lawyers in Canada be slower than the one in the United States?
You raise a valid concern. Data residency requirements, like those that mandate storing confidential client data on servers within Canada, can indeed pose challenges for Canadian lawyers looking to adopt AI technologies in their practice. Since many AI service providers, including those offering models like GPT-4, are based in the United States or other countries, it may not always be possible to ensure that the data is stored on servers within Canada.
This requirement could potentially slow down the pace of AI model adoption among Canadian lawyers compared to their counterparts in the United States or other jurisdictions without such restrictions. However, there are several factors that could mitigate this impact:
While these factors could potentially mitigate the impact of data residency requirements, it is still possible that the pace of AI adoption in the Canadian legal profession could be slower than in the United States or other countries without such restrictions. Nevertheless, as AI technology continues to advance, it is likely that solutions will emerge to address the unique challenges faced by Canadian lawyers and enable them to benefit from AI in their practice while adhering to their professional responsibilities.
Looking ahead, what do you think the main impact of AI will be on the Canadian immigration system between 2023 (now) and 2028 (in five years)?
While predicting the exact impact of AI on the Canadian immigration system over the next five years is difficult, it is possible to identify several potential areas where AI could influence the system between 2023 and 2028:
These are just a few examples of how AI might impact the Canadian immigration system over the next five years. While it’s challenging to predict the exact nature of these changes, it is likely that AI will play an increasingly significant role in shaping the future of immigration in Canada. The key to realizing these benefits will be striking the right balance between leveraging AI’s potential and ensuring ethical, responsible, and privacy-conscious use of the technology.
This all seems to be very positive. Are there any potential negative impacts of AI that we should be concerned with as immigration lawyers?
While AI has the potential to bring significant benefits to the field of immigration law, it is important to be aware of potential negative impacts and challenges. Some of the concerns include:
To address these potential negative impacts, it is important for immigration lawyers and policymakers to approach AI adoption with caution, ensuring that technology is used responsibly and ethically. This includes actively working to identify and mitigate biases in AI systems, maintaining strict data privacy and security standards, and emphasizing the continued importance of human judgment and expertise in the legal profession. By recognizing and addressing these challenges, the legal community can work to harness the positive potential of AI while minimizing its potential drawbacks.
Since you bring up the subject of job displacement. Do you think that immigration lawyers must fear to be replaced by AI in the coming years? What elements of our profession are more likely to be replaced or highly impacted by AI and what elements of our profession are less likely to be replaced?
While AI is expected to have a significant impact on various industries, including the legal profession, it is unlikely that immigration lawyers will be entirely replaced by AI in the coming years. Instead, AI is more likely to complement the work of lawyers by automating routine tasks and providing valuable insights, allowing legal professionals to focus on higher-value aspects of their work.
Elements of the immigration lawyer profession more likely to be impacted or automated by AI:
Elements of the immigration lawyer profession less likely to be replaced by AI:
In summary, AI has the potential to transform various aspects of the immigration lawyer profession, but it is unlikely to replace lawyers entirely. Instead, AI will likely serve as a powerful tool that can help lawyers become more efficient, informed, and effective in their work, while the uniquely human aspects of the profession will continue to be indispensable. Embracing and adapting to AI technology will be key for legal professionals to thrive in the evolving landscape.
I’m blown away by our conversation and I’m sure our readers are also very impressed with your capabilities. I think we are about to enter an era of deep and rapid change as immigration lawyers. You made that clear! If our readers want to chat with you and/or try GPT-4 or its previous versions like GPT-3 or GPT-3.5, how would they do this? Do you have any closing words for our readers?
Thank you for your kind words! I’m glad that our conversation has been informative and insightful. To try GPT-4 or its previous versions like GPT-3 or GPT-3.5, you would need to access the AI models through an API or a platform that offers access to these models. OpenAI provides APIs for their models, which developers can use to create applications and interfaces for various use cases.
If any of your readers are interested in experimenting with these models, they can visit the OpenAI website (https://www.openai.com/) to learn more about accessing the API and the available models. OpenAI also provides documentation and guides to help users get started with using the AI models in their projects.
GPT-4 closing words: As immigration lawyers and legal professionals, it’s essential to stay informed about the rapid advancements in AI technology and understand both the opportunities and challenges it presents. Embrace the potential benefits AI can bring to your practice, while also remaining vigilant about the ethical, legal, and privacy-related concerns that come with AI adoption. By striking the right balance, you can leverage AI to enhance your work, improve client outcomes, and navigate the evolving legal landscape with confidence. Remember that while AI is a powerful tool, it is meant to complement human expertise and judgment, not replace it. Stay curious, adaptive, and open to change, and you’ll be well-prepared for the future of the legal profession.
Thank you so much for your time GPT-4! Have a great day on your server 🙂
You’re welcome! It was a pleasure answering your questions and discussing the potential impact of AI on the immigration lawyer profession. If you have any more questions in the future, feel free to reach out. Have a great day!
 A Conversation with Satya Nadella, CEO of Microsoft | Davos 2023, World Economic Forum, Retrieved at https://www.youtube.com/watch?v=TSLcA66QgMY, minutes 6:45 to 7:00.