Interview with Corporate Lawyer Mikkeli Han - Discussion about AI regulation & law
Discussed the transformation of the legal market, AI integration in legal work, AI policies, and its structural challenges to be resolved
INTERVIEWS
YAFI Interview team
5/29/2026
Interview held on May 25th, 2026
AI, Law, and Human Responsibility: Lessons from an Interview with Mr. Han
In a recent interview with the Youth AI Future Institute, Mr. Han, an M&A lawyer working on cross-border transactions across regions such as Korea, Singapore, and the broader APAC market, offered a realistic and cautious perspective on the growing role of artificial intelligence in the legal industry. Rather than treating AI as either a revolutionary replacement for lawyers or a harmless productivity tool, Mr. Han presented a more balanced view: AI is already changing legal work, but the final responsibility, judgment, and ethical burden must remain with human lawyers.
How AI Is Already Entering Legal Practice
The interview began by asking Mr. Han about his background and how AI has started to intersect with his work as an M&A lawyer. He explained that M&A work often involves identifying transactional risks, reviewing deal terms, and managing complex cross-border legal issues. In this context, AI is increasingly useful for document review, risk identification, and drafting. These are tasks that require processing large amounts of information, finding relevant issues, and producing initial written outputs.
One of Mr. Han’s key insights was that AI is not merely a future possibility in law; it is already being actively used, especially for work traditionally assigned to junior lawyers. For example, AI can help generate first drafts, summarize documents, identify issues, and support legal research. In this sense, AI is already becoming a partial substitute for some lower-level legal tasks. However, he also emphasized that law firms still need to know how to use AI properly. The tool itself does not automatically produce reliable legal judgment; its value depends on how carefully and professionally it is used.
Hype Versus Reality in Legal AI
The interview then moved to the question of whether the current excitement around AI in law matches reality. Many people argue that law is one of the professions most likely to be replaced by AI because legal work often involves language, documents, precedent, and reasoning. Mr. Han’s answer suggested that this view is partly true but incomplete.
He acknowledged that AI can already perform certain legal tasks quickly and efficiently. In fact, he noted that many firms are already using AI for junior-level work and first-draft outputs. At the same time, this does not mean that lawyers themselves are becoming unnecessary. The legal profession involves more than producing text. It requires understanding context, anticipating consequences, negotiating risks, interpreting human intentions, and making professional judgments. Therefore, the reality is not a simple replacement, but a shift in the kind of work lawyers are expected to do.
The lesson here is that AI may reduce the value of repetitive legal work while increasing the importance of higher-level human judgment. Future lawyers may not be valued simply for drafting basic documents, but for reviewing, correcting, interpreting, and taking responsibility for AI-assisted outputs.
Responsibility When AI Makes Mistakes
A central ethical question in the interview was who should be responsible if an AI system helps draft a legal argument or research case precedents, but the result later turns out to be wrong or fabricated. Mr. Han’s answer was clear: the lawyer must remain responsible. Professional judgment cannot be outsourced.
This point is especially important because AI systems can produce convincing but false information. In law, such errors can have serious consequences, including misleading a court, damaging a client’s case, or weakening public trust in the legal system. Mr. Han emphasized that AI is still only a tool. Even if AI helps produce a legal argument, the lawyer is the professional who submits, relies on, or acts upon that argument.
The broader lesson is that AI does not remove human responsibility. Instead, it may increase the need for careful human review. A lawyer using AI must treat its output like the work of a junior assistant: useful, but never automatically trustworthy. Just as a senior lawyer would check a junior lawyer’s draft, AI-generated work must be verified before it is used in any professional setting.
AI Judges and the Human Nature of Justice
The interview also asked whether AI judges could ever be ethically permissible. If AI can analyze statutes and precedent, some may argue that it could make legal decisions more consistently than humans. However, Mr. Han expressed ethical concern about this idea. In his view, justice requires human judgment, empathy, and contextual understanding. These human layers may be lost if judicial decisions are delegated to machines.
This response highlights an important distinction between legal calculation and justice. Law is not only about applying rules mechanically. It also involves fairness, intention, proportionality, and an understanding of human circumstances. A judge must consider not only what the law says, but how it applies to real people in complex situations. AI may be able to support legal research or organize precedent, but it cannot fully replace the moral and human dimension of judging.
The lesson is that some parts of the legal system may be improved by AI, but not all parts should be automated. Especially when a decision directly affects a person’s rights, freedom, punishment, or dignity, human judgment remains essential.
Preserving Human Values in an AI-Assisted Legal System
Another major topic was how human values such as fairness, intention, rights, and proportionality can be preserved as AI becomes more involved in law. Mr. Han explained that law is grounded in human judgment, but he also recognized that human judgment itself is not perfect. Humans can make mistakes, misunderstand context, or act with bias. Therefore, AI is not necessarily negative. It could help reduce certain human errors if it is properly designed and carefully guided.
However, Mr. Han stressed that context and interpretation remain crucial. AI must be trained and evaluated not only to process legal information, but also to recognize context and reduce bias. Human review is essential because legal decisions often depend on details that are not purely technical. A legal system that uses AI responsibly must therefore combine technological assistance with strong human oversight.
This creates an important lesson: the goal should not be to choose between humans and AI, but to design a system where AI improves efficiency while humans preserve accountability, ethics, and judgment.
Where Human Sign-Off Must Remain Mandatory
When asked which stages of the legal process should require mandatory human participation regardless of AI’s future capabilities, Mr. Han again emphasized human responsibility. He argued that human sign-off must remain mandatory for court submissions, contracts, sentencing, and judicial decisions. These are areas where legal consequences are too serious to leave entirely to automated systems.
His answer suggests that the most important legal threshold is not whether AI can assist, but whether AI should be allowed to decide or finalize. Drafting assistance may be acceptable. Research assistance may be useful. Document review may become more efficient. But final approval, especially in matters affecting legal rights, obligations, punishment, or business risk, must remain with a human professional.
The lesson is that AI can participate in the legal process, but it should not become the final authority. Final responsibility must remain traceable to a human being.
Why AI Law and Regulation Move Slowly
The final part of the interview focused on why AI governance and legislation move slowly compared to technological development. Mr. Han explained that regulation is inherently reactive. Laws often emerge after risks have become visible or proven. Legislators are cautious because legal rules are difficult to reverse once created, and regulating too early may unintentionally restrict innovation.
He also pointed out that businesses and technology companies move according to different incentives. Companies often want to capture market share quickly, while legal systems move more carefully. This creates a gap between the speed of innovation and the speed of regulation. In cross-border transactions, this problem becomes even more complicated because different countries have different data policies, legal standards, and attitudes toward AI.
The lesson from this part of the interview is that slow regulation is not only a failure of the government. It reflects a deeper tension between innovation, risk, business incentives, and legal caution. AI is not just a single product that can be regulated easily; it is a broad capability that can be applied across industries. This makes it difficult for legal systems to respond quickly without either falling behind or overregulating too early.
Conclusion: Trust, Caution, and the Future of Legal AI
Perhaps the clearest summary of Mr. Han’s view came when he was asked whether he personally trusts AI. His answer was direct: not completely. He is willing to use AI for first drafts, issue lists, and initial support, but he remains cautious and always double-checks its work. This attitude captures the central message of the interview.
AI is already transforming the legal profession, especially by improving efficiency in drafting, review, and research. However, Mr. Han’s perspective shows that the future of law cannot be reduced to automation. The legal profession depends on trust, responsibility, human judgment, and ethical accountability. AI may become a powerful assistant, but it should not become an unreviewed authority. In law, the most important question is not only what AI can do, but who remains responsible when its outputs affect real people, real rights, and real consequences.