Pacific Meta Develops 'AutoFund,' an AI Agent for Autonomous Asset Management, and Begins Internal Test Operations

Pacific Meta Inc. (Minato-ku, Tokyo; CEO: Shota Iwasaki) has developed "AutoFund," an asset management product that utilizes AI agents to autonomously execute transactions on the blockchain, and has begun test operations using its own capital.

The company is currently exploring potential collaborations with financial institutions and asset management firms to address regulatory, auditing, and operational design considerations related to this initiative.

[Note] This announcement is not intended to solicit investments or provide investment advice. Legal and regulatory issues regarding this initiative are being reviewed and organized in consultation with external experts.

---

■ Overview of AutoFund

AutoFund is an asset management product where AI references on-chain data to perform end-to-end rebalancing decisions and transaction execution based on market conditions. During the test phase, the company is allocating its own assets to AutoFund, allowing the AI agent to execute asset rebalancing based on real-time on-chain data.

The system is designed to limit AI execution within pre-defined parameters, including target assets, conditions, and transaction limits. This allows the company to verify whether controlled, automated operations are feasible. Future service models will be determined based on the progress of regulatory and legal assessments.

---

■ Why 'AI × Blockchain' is Crucial for Asset Management

1. Why Blockchain is Necessary for AI Asset Management

On the blockchain, AI can move assets directly through code. While traditional finance requires human intervention via banks or brokerages, blockchain removes these constraints, enabling the automation of both decision-making and execution. Furthermore, all transactions are recorded in an immutable format, allowing anyone to verify the AI's operational behavior. AutoFund utilizes blockchain to integrate "autonomous AI execution," "transparency," and "controlled authority."

2. Expanding On-Chain Data Continuously Improves AI Accuracy

All transaction data on the blockchain is public and structured for direct AI consumption. As on-chain adoption grows, the volume of data available for AI learning increases, enhancing the accuracy of operational decisions. Unlike traditional financial markets where data is siloed within internal systems, blockchain data is globally accessible, providing a structural advantage for AI-driven operations.

3. Humans Cannot Monitor 24/7 Markets Alone

On-chain financial markets operate continuously worldwide. It is physically impossible for humans to monitor hundreds of investment targets simultaneously and move assets at optimal times. Autonomous monitoring, decision-making, and execution by AI agents have the potential to democratize operational capabilities previously reserved for specialized teams at large financial institutions.

---

■ Technical Architecture of AutoFund

Safe (Smart Contract Wallet)

A wallet used to store and move assets for test operations.

Allowance Module (Safety Mechanism)

Sets on-chain limits for operations within a specific period, ensuring transactions are executed only within permitted ranges.

AI Agent (Operator)

The AI agent...