rewrite this title Sakana AI And MUFG Enter $34M Agreement To Automate Banking Document Generation

by
Alisa Davidson

Published: June 16, 2025 at 9:31 am Updated: June 16, 2025 at 9:31 am

by Ana

Edited and fact-checked:
June 16, 2025 at 9:31 am

To improve your local-language experience, sometimes we employ an auto-translation plugin. Please note auto-translation may not be accurate, so read original article for precise information.

Japan-based AI company Sakana AI disclosed that it had finalized a $34 million agreement with Mitsubishi UFJ Financial Group (MUFG) to implement AI-driven automation in banking document creation, including materials such as credit approval memos. 

As outlined in a joint statement, the agreement will span a period of more than three years. A six-month pilot program is scheduled to commence in July, during which MUFG’s banking unit will utilize Sakana AI’s system–The AI Scientist. This system, originally developed for automating tasks in scientific research—such as manuscript drafting and peer review—will now be adapted for financial document generation. According to the companies, the collaboration includes the continued involvement of Ren Ito, Sakana AI’s cofounder and chief operating officer, who will serve as an AI adviser to MUFG Bank throughout the duration of the project. 

Ren Ito noted that while the AI system is expected to accelerate the production of credit memos, broader decisions related to lending will remain complex and not solely determined by AI tools. He also acknowledged that the partnership will need to address long-term considerations, particularly those extending beyond basic gains in operational efficiency.

Sakana AI, a company established in 2023 by David Ha, has emerged as Japan’s fastest-growing unlisted startup to attain a valuation of $1 billion. The firm has secured approximately $208 million in funding from a diverse group of investors that includes MUFG Bank, Sumitomo Mitsui Banking Corporation, Mizuho Financial Group, and Nvidia. 

Sakana AI’s The AI Scientist is an open-source system engineered to manage the complete research lifecycle without human input: it brainstorms novel scientific ideas, conducts automated experiments by writing and executing code, compiles and visualizes data, drafts full academic papers, and performs its own peer review. 

Meanwhile, the firm is also known for its research approach inspired by patterns found in natural systems, which it applies to the development of AI technologies. Among its recent innovations is the Darwin Gödel Machine, an AI agent engineered to autonomously revise and enhance its own code. 

In addition, the company has introduced a novel method called Text-to-LoRA, a hypernetwork-based system designed to create task-specific Low-Rank Adaptation (LoRA) modules for large language models. This method enables the generation of these modules directly from textual task instructions, representing a step forward in the customization and adaptability of LLMs.

About The Author

Alisa, a dedicated journalist at the MPost, specializes in cryptocurrency, zero-knowledge proofs, investments, and the expansive realm of Web3. With a keen eye for emerging trends and technologies, she delivers comprehensive coverage to inform and engage readers in the ever-evolving landscape of digital finance.

More articles

Alisa Davidson

Alisa, a dedicated journalist at the MPost, specializes in cryptocurrency, zero-knowledge proofs, investments, and the expansive realm of Web3. With a keen eye for emerging trends and technologies, she delivers comprehensive coverage to inform and engage readers in the ever-evolving landscape of digital finance.

More articles