neoAI Inc. (Headquarters: Minato-ku, Tokyo; CEO: Shunsuke Chiba; hereinafter referred to as "neoAI") is pleased to announce that it has implemented an advanced audit functionality in its generative AI platform, "neoAI Chat," to ensure the safe and appropriate utilization of generative AI. Ikedasenshu Holdings Inc. (Headquarters: Osaka, Osaka Prefecture; President & CEO: Hirokazu Sakaguchi; hereinafter referred to as "Ikedasenshu Holdings") has decided to adopt this functionality. [About neoAI Chat Advanced Audit Functionality] The "neoAI Chat Advanced Audit Functionality" is an optional feature for "neoAI Chat" that provides guardrail and audit log acquisition functions for corporations. It supports internal controls, incident investigations, and audit responses through a guardrail function that monitors user input and prevents risky submissions, and an audit log acquisition function that records system authentication and operations within the system. [Initiatives at Ikedasenshu Holdings] 1. Background and Objectives As the full-scale use of "neoAI Chat" by companies progressed, neoAI recognized the growing need for enhanced governance of generative AI. In particular, companies with high security standards, such as financial institutions, required controls for the risk of users unintentionally inputting information that does not comply with policies, and the acquisition of evidence to regularly verify safe usage. To address these challenges, neoAI, together with Ikedasenshu Holdings, began efforts to build an advanced audit functionality that could meet the company's standards. 2. Implementation Details and Results neoAI conducted a Proof of Concept (PoC) with Ikedasenshu Holdings, discussing the required audit methods, and refining the detection targets and operational rules for the guardrail function in the context of banking operations, as well as improving detection accuracy. The PoC primarily verified the following three aspects: (1) Automatic Detection During