PRONEXUS Group AI Usage Policy

「プロネクサスAI利用ポリシー」は、「プロネクサスグループAI倫理ポリシー」を具体化した「グループ共通の標準指針」として定めたものです。お客様との個別のご契約または合意がある場合には、そちらの定めが優先されます。当社グループは、技術と社会の変化に応じて、本ポリシーの必要な見直しを行っていきます。

As well as setting out specific usage conditions and safe management standards for AI (LLM, etc.) for the operations of the Group (operations outsourced to it and services offered by it) in accordance with the PRONEXUS Group AI Ethics Policy, this policy sets out standardized operating guidelines for the use of AI in each of the Group’s services.

(1) Security first
Before use, we confirm that the environment is one in which data entered is not used for training.

(2) Human-in-the-loop approach
AI-generated output is strictly supplementary in nature, and humans take responsibility for verifying quality, accuracy, and whether any rights have been infringed.

(1) Relation to contracts
This policy does not constitute or amend the content of individual contracts and terms of use with clients, but defines the technical specifications used in the execution of operations, and internal quality standards.

(2) Individual agreements take priority
In cases where, in a client contract, the use of AI is forbidden, or specific usage conditions are agreed separately, the provisions of that contract take priority.

In cases where the Company makes use of AI (LLM) for services outsourced to it or services that it offers, it uses one of the following environments to ensure that data entered is not used for training the models of AI vendors.

(1) Cloud LLM environment
Cloud services (within our tenants, etc.) in which the contract with the AI vendor explicitly states that entered data will not be used for the training of models.

(2) Local LLM environment
Local LLM environment built and operated within our data center, etc.

(1) Statement or explanation
In cases where AI was used during the creation of content, depending on the nature of the deliverables and the contract, etc. with the client, there will be an appropriate statement or explanation to the effect that AI has been used.

(2) Guarantee of accuracy and adherence to the law
Humans perform fact-checking (verification and confirmation) to ensure that generated output is accurate and does not infringe the intellectual property rights of third parties (adherence to the law with regard to copyrights, etc.).

(3) Transparency as an operator of generative AI business
In cases where the Company is the developer or provider of generative AI, it strives to gain the understanding and trust of user by appropriately disclosing and explaining mechanisms related to the use of training data and protection of generated output rights (technical mechanisms, etc. to ensure verifiability and transparency of data protection, and prevent infringement of rights, within a reasonable scope).

(4) Good-faith engagement
In cases where disadvantages for the client cannot be avoided despite the above-mentioned verification being implemented, the Company will cooperate in good faith, and strive to resolve and mitigate the situation. The Company will follow the scope and content of specific responsibilities of the Company set out in individual contracts and other agreements with the client.

(1) Visualization of usage
Each department specifies the AI tools that it uses and the situations in which they are used, and manages them appropriately in accordance with the list (AI verification and usage, etc.) specified by the Company.

(2) Continuous improvement
This policy will be reviewed periodically in response to technological innovation, requests from clients, and changes in the social situation.

Established in October 2026

Takeshi Ueno, Representative Director,
PRONEXUS INC.