my timesThe Korea Times

Lawyers, counselors grapple with what AI can and cannot replace

Listen

AI automates routine tasks, but professionals say machines cannot replicate human expertise

Editor’s note

As Korea pushes to embed artificial intelligence (AI) and robotics across its factories and offices, the impact on jobs is no longer abstract. Workers are already feeling it. This is the third in a four-part series examining how AI is reshaping work on the ground — the opportunities it creates, the protections it erodes and the rules that unions, employers and policymakers are, or are not, putting in place to govern that transition. — ED.

A Seoul-based lawyer surnamed Ha spends about 600,000 won ($410) a month on artificial intelligence (AI) subscriptions such as Claude and Korean legal platforms such as SuperLawyer and LBOX. Hiring a junior lawyer would cost him close to 100 million won a year, roughly 14 times as much. The math, he says, is no longer close.

“Without AI, I’d probably have one more person in the office by now,” Ha told The Korea Times, adding that over the past two years, he has used large language models intensively for work. “I still need to review the quality of final drafts, but the need to hire new lawyers has definitely gone down.”

Only a few years ago, Korean law firms routinely hired swarms of new law school graduates to plow through ruling precedents, check statutes and draft the first versions of briefs. Now, more managing partners reach for a subscription instead of a job posting, and the same calculation is playing out across other white-collar professions, from legal offices to counseling rooms, as AI forces a reckoning over what kind of work still requires a human.

“I don’t regret it at all. In the past, all of this would have gone toward labor costs,” Ha said.

In his small office, the shift has been quiet but profound. Ha built his own specialized system by scanning his legal reference books and feeding them into a personalized database. Now, he can ask questions in everyday language and the AI searches his digital library, identifies the most relevant sections and provides analysis — a process he says he has fully automated.

It is unclear how widespread such AI adoption is. But according to LBOX, the most popular legal AI platform among Korean lawyers, it has roughly 22,000 members; there are a total of around 38,000 registered lawyers here.

Junior lawyer problem

According to Ko Hak-soo, a Seoul National University law professor who specializes in AI and technology law, the legal establishment is dangerously unprepared for what’s coming.

“The most troubling thing is that no one is systematically examining what changes are taking place with the rise of AI,” the professor said. “Everyone has opinions about changes brought by AI, but there is no organized effort to assess the real impact or prepare for it, especially in terms of training aspiring lawyers or those who recently became lawyers.”

Ko believes that AI adoption, still in early stages, threatens to deepen polarization within the profession. He described a scenario where experienced lawyers at major firms will capture even more value using AI, while tasks traditionally handled by junior lawyers are automated.

“In the United States and Europe, bar associations and law schools are already debating this ‘AI pyramid’ effect — a narrowing base of younger junior lawyers and an expanding layer of senior partners,” he said. “The tasks that junior associates used to do are the ones AI can more easily replace. So fewer junior lawyers will get real field experience, and without that experience, they could struggle to advance in the profession.”

To stay relevant, legal education needs to evolve, with simulation-based learning and mock scenarios potentially substituting for disappearing real-world training opportunities, he said.

“This is the time to seriously consider a major shift in how we train legal professionals,” said Ko. “But instead, many skip all the important questions and jump straight to debates about whether we have too many lawyers. I think that’s the wrong question.”

A man speaks to an artificial intelligence-powered counselor during Seoul Smart Life Week, an annual global tech exhibition, at Coex in southern Seoul, Sept. 30, 2025. Newsis

A man speaks to an artificial intelligence-powered counselor during Seoul Smart Life Week, an annual global tech exhibition, at Coex in southern Seoul, Sept. 30, 2025. Newsis

Human encounters

Meanwhile, another profession built on human connection is navigating new technology with more confidence.

Kim Jang-hoi, head of the Korean Counseling Association, said psychological counselors face little immediate threat from AI chatbots — though he acknowledged that AI is creeping into how large organizations deliver mental health support.

“Individual counseling centers haven’t seen much change yet,” said Kim, whose association represents over 40,000 members. “But for larger institutions and state agencies, AI chatbots are being seriously discussed because it comes down to budget.”

Earlier this year, the Seoul Metropolitan Government launched “Maeum-e,” an AI-powered counseling chatbot, as part of efforts to combat rising mental health challenges and provide 24-hour suicide prevention support. Commercial apps like BlueSignum’s Lighthouse also use AI to analyze users’ emotions and recommend personalized self-care activities.

A survey released in March by Gyeonggi Research Institute found that 27 percent of Seoul metro area residents with normal mental health use AI counseling services, with the number increasing to 53 percent among those with severe depression. Researchers said the findings suggest that people in more severe distress may avoid in-person counseling, making anonymous AI tools an appealing alternative despite their limitations.

Yet Kim dismissed concerns that AI would displace human counselors. Chatbots can provide information, but they fundamentally cannot replicate what human experts provide: genuine human encounters.

“Counseling, fundamentally, isn’t about getting answers — it is about the nuances of face-to-face interaction, even the meaning found in moments of silence,” Kim said.

That distinction matters as AI becomes more sophisticated. In several countries, regulators and researchers have examined cases where users died by suicide after chatbots appeared to echo or reinforce self-harm ideation instead of intervening or flagging risk.

Such tragedies highlight why counseling requires human judgment and ethical responsibility that algorithms cannot provide. However, Kim said, this doesn’t mean rejecting AI entirely.

Rather than viewing AI as a threat, many counselors are adopting it as a tool, using it to transcribe session recordings or analyze therapeutic interactions.

Despite institutional budget pressures and AI’s growing presence, Kim believes demand for human counselors will increase as technology becomes more pervasive.

“In a world where AI becomes ubiquitous, human encounters become even more valuable,” he said. “Counseling is a profession that helps people through real human connection — and I believe demand for that will only grow.”