[Edaily Marketin Reporter Song Seung-Hyeon ] In an era where work automation has become the norm as agentic artificial intelligence (AI) is rapidly integrated into various office tasks, there are now cases where AI is handling mergers and acquisitions (M&A)—a process that encompasses due diligence, legal matters, and more. Specifically, a case has emerged in the U.S. where the entire M&A process—from negotiating terms to drafting the definitive agreement and closing the deal—was handled using an AI platform.

According to the investment banking (IB) industry on the 16th, Soxton, an AI legal startup based in New York, announced in March that it had acquired Cipher, a security firm specializing in agentic applications. At the time, Soxton stated in a press release that the acquisition of Cipher was completed entirely on the Soxton platform. The company further elaborated that every stage of the transaction—from negotiating terms to drafting the main contract and closing the deal—was handled using its AI platform, resulting in savings of $80,000 (approximately 110 million won) in legal fees.

This is not an M&A deal conducted without a law firm; rather, it is an instance where an AI platform—which bills itself as a law firm—handled its own deal using its own platform. Soton was founded by Logan Brown, a Harvard Law School J.D. graduate who previously worked as a corporate attorney at the major U.S. law firm Cooley. The company raised $2.5 million in a pre-seed round led by Moxy Ventures last December. Its core service automates everything from incorporation to equity management, fundraising, and compliance for early-stage startups.

In an interview with Forbes last August, Brown explained the company’s service, stating, “While drafting, redlining, and contract review begin with AI trained on our templates, an experienced attorney reviews the AI-drafted contract before it is delivered to the client.” The company operates on a flat-rate model: for $20 per month, customers receive contract templates, and for $100, they receive a customized contract reviewed by a lawyer within four hours. However, the lawyer review process described by Brown is based on client work; it has not been disclosed whether a lawyer reviewed the acquisition agreement for Cypher, a deal involving the company itself.

Sutton’s acquisition of Cypher is also drawing attention within the domestic M&A legal advisory industry. An M&A specialist at a major law firm explained, “The extent to which AI can replace the work of lawyers is a matter of great interest,” adding, “Given the predictions that AI could replace a certain portion of M&A-related work, the industry is watching this U.S. case with great interest.”

Similar initiatives are also underway in Korea. “Listing,” the M&A platform developed by DeepSearch, an AI company specializing in corporate data, recorded 2,500 subscribers, 1,400 cumulative company reviews, 379 ongoing transactions, and 4 completed deals just nine months after its April 2024 launch. These figures are as of early last year. In particular, Listing positions the fact that the M&A brokerage process is conducted by AI as its key differentiator.

Another example is “CookieDeal” by Fractal Technology, a company that emerged from the third cohort of Antler Korea’s Startup Generator. CookieDeal received 700 million won in seed funding from Spring Camp and Antler in July 2024. It targets non-audited corporations with a valuation between 1 billion and 10 billion won, and the company’s vision is to “create an AI-based investment bank for small and medium-sized enterprises by combining technology with finance.” CookieDeal was also selected as a new M&A advisory firm by the Korea Venture Capital Association (VC Association).

However, the scope of operations for domestic platforms generally stops at deal sourcing, matching, and due diligence. The drafting of contracts and the closing stages are handed over to advisory firms and law firms. A prime example is Listing, which entrusted the structuring and legal advisory services for its cross-border deal with Japan to the law firm DLJ. This is less a matter of technological capability and more a systemic issue. The Attorney Act prohibits non-lawyers from handling legal matters, and due diligence and valuation overlap with the scope of the Certified Public Accountant Act. The moment a platform drafts a contract and pushes the deal through to closing, it risks accusations of unqualified legal practice.

For this reason, AI-driven M&A has yet to take root in South Korea. Another attorney specializing in M&A at a major law firm stated, “Repetitive tasks, such as organizing due diligence materials or drafting contracts, have already been largely handed over to tools,” adding, “What remains is determining how to structure the deal and which clauses to compromise on, and the entity responsible for those decisions is still the lawyer.” He continued, “However, there is also a sense of crisis that even these tasks could be replaced as AI becomes more sophisticated.”