Yi Joon-ki, managing accomplice of Bae, Kim & Lee (BKL), speaks throughout a current interview with The Korea Occasions at his workplace in Seoul. Courtesy of BKL
As synthetic intelligence (AI) turns into an ordinary software throughout the authorized occupation and past, purchasers will finally decide regulation corporations not by whether or not or how a lot they use the know-how, however by the standard of the outcomes they ship with it, stated Yi Joon-ki, managing accomplice of Bae, Kim & Lee (BKL).
“Shoppers do not likely care what instruments we use,” Yi stated in a current interview with The Korea Occasions at his workplace. “What they need is a compelling and helpful evaluation — one which addresses their wants. In the long run, it’s BKL’s judgment, and BKL shall be liable for it. So long as we ship a dependable reply, that’s what issues to them.”
The Seoul-based regulation agency has emerged as probably the most lively adopters of generative AI amongst Korean regulation corporations, rolling out the know-how throughout the group this yr after conducting pilot initiatives in litigation, authorized analysis, due diligence, doc drafting and different authorized duties.
Yi stated BKL’s broader ambition just isn’t merely to automate routine work, however to revamp authorized workflows in order that legal professionals can ship deeper evaluation, sharper strategic judgment and extra helpful recommendation to purchasers.
“The introduction of AI in work can solely elevate the ground,” Yi stated. “However what issues extra is elevating the ceiling, which relies on how it’s utilized by individuals … The true competitors shall be over who can use it to provide a greater last product for the consumer.”
The excellence is central to BKL’s method. AI can shortly summarize paperwork, determine clauses, evaluate testimonies and draft routine paperwork — duties that beforehand required many hours of labor by junior legal professionals. However Yi stated the better aggressive benefit will come from combining the know-how with a agency’s accrued know-how, the judgment of skilled legal professionals and methods designed to make the perfect use of AI-generated supplies.
“If a regulation agency merely asks AI a query and accepts one reply, that’s one stage,” he stated. “But when it will probably generate a number of attainable approaches, evaluate them, confirm the information and authorized grounds, and use them to construct a greater technique, it turns into a very totally different sort of service.”
When ChatGPT was launched to the general public in late 2022, Yi thought it was untimely for BKL to undertake the know-how, which appeared promising however not reliable sufficient for work. Safety, notably consumer confidentiality, remained a key concern.
However as AI instruments improved, he concluded that each the know-how and its inner governance safeguards had superior sufficient to handle the dangers. Yi additionally pointed to altering company attitudes towards “shadow AI” — staff’ use of unapproved AI instruments outdoors firm methods — even at main firms with stringent safety necessities. Moderately than attempting to ban AI use altogether, he stated, it might be higher to make use of managed platforms that might carry such exercise below institutional management.
Requested the place AI has proved most helpful, Yi pointed to issues involving massive volumes of paperwork, together with complicated prison circumstances.
Attorneys, for instance, as soon as needed to spend many months reviewing interrogation data and witness statements to determine inconsistencies, contradictions and corroborating proof — a job AI can now deal with in a fraction of the time.
AI can be altering the best way BKL handles cross-border issues, permitting its legal professionals to hold out extra substantive preliminary analysis.
“As an alternative of asking abroad counsel an open-ended query about what the regulation is there, we are able to now say, ‘Based mostly on our analysis, we expect the reply is sure. Are you able to affirm that?’ That adjustments the character of cross-border authorized work fully,” Yi stated.
Yi stated BKL’s subsequent aim is to attach AI with the agency’s 4 a long time of accrued authorized supplies and the tacit judgment of its skilled legal professionals. Moderately than limiting an inner AI system to a slim set of current paperwork in a single observe space, the agency has used AI-aided preprocessing to arrange its most up-to-date years of data throughout the agency and goals to finish the broader historic archive by year-end. The brand new system will incorporate BKL legal professionals’ reasoning standards, permitting it to provide top quality recommendation that can’t be imitated by general-purpose AI instruments, he stated.
BKL recorded 440.2 billion received ($320 million) in income final yr, overtaking Lee & Ko for the primary time in six years to change into Korea’s second-largest regulation agency by income after Kim & Chang. Its development has continued this yr, with income reported in value-added tax filings for the primary half rising 29.3 % from a yr earlier.
Yi outlined three priorities for BKL’s subsequent stage: increasing AI throughout day-to-day authorized work to ship tangible worth to purchasers; constructing an outcome-driven tradition and creating expertise able to delivering outcomes for purchasers; and strengthening the agency’s skill to resolve complicated cross-border issues.
“These duties finally converge on one aim: qualitative development and stronger worldwide competitiveness,” Yi stated. “What issues just isn’t merely changing into bigger, however elevating the extent at which we clear up tough issues — and making certain that stage meets worldwide, not simply home, requirements.”
