Generative AI usage guidelines

1 Purpose of these guidelines

This guideline explains the things that OpenFashion Inc. should be aware of when using generative AI such as MaisonAI, ChatGPT, and Midjourney in its business operations.

Generative AI is useful for improving work efficiency and generating new ideas, but depending on the content of input data and how the generated products are used, it may violate laws or infringe on the rights of others. Please read these guidelines carefully and use generation AI effectively.

2 Generation AI covered by this guideline

The generation AIs targeted by this guideline are MaisonAI, ChatGPT, and Midjourney. If you wish to use other generation AI, please contact the corporate department.

3 Applications where the use of generated AI is prohibited

OpenFashion prohibits the use of generated AI for the following purposes and tasks.

(1) Analysis and utilization of customer personal information obtained through our own services and client operations

(2) Analysis and utilization of confidential information obtained in client business]

4 Structure of this guideline

Generative AI basically has a structure in which the user inputs some data, some processing (storage, analysis, generation, learning, re-provision, etc.) is performed, and the result (product) is obtained. is.

Therefore, this guideline consists of the following two parts.

• Points to note when entering data
• Points to note when using the product

5 Points to note when entering data

The data to be input (sent) to the generation AI includes a wide variety of types, but from the perspective of the necessity of processing intellectual property rights and compliance with laws and regulations, the following types of data should be input in particular: Caution is required.

(1) Data whose copyright is owned by a third party (texts created by others, etc.)

In principle, the act of simply inputting another person's copyrighted work into a generating AI does not constitute copyright infringement.

However, if the purpose is to generate an AI product that is the same or similar to another person's copyrighted work, the act of input itself may constitute copyright infringement.

Also, if the generated data is the same or similar to the input data or existing data (copyrighted work), please be aware that the use of the generated data may infringe on the copyright of the copyrighted work. please. Specifically, please refer to the section ``6(2) The act of using the product may infringe on someone's existing rights.''

In addition, the creation of original models through fine-tuning and the use of other people's copyrighted works for so-called prompt engineering are, in principle, not considered to constitute copyright infringement.

(2) Registered trademarks/designs (logos and designs)

Inputting logos, designs, etc. that are registered as trademarks or designs into generation AI does not constitute trademark infringement or design rights infringement.

However, it should be noted that, like copyrighted works, this point only concerns the "act of input." The act of commercially using a trademark or design that is the same or similar to another person's registered trademark or design, whether intentionally or accidentally, constitutes trademark infringement or design rights infringement.

In other words, there is little need to research registered trademarks and designs when inputting logos and designs to generated AI, but research is necessary when using generated products.

(3) Photographs and names of celebrities

The act of inputting a celebrity's face photo or name into a generation AI does not constitute a violation of the publicity rights held by the celebrity.

However, please be aware that commercial use of names, portraits, etc. of celebrities generated using generated AI constitutes a violation of publicity rights.

(4) Personal information

In ChatGPT, the entered data will be used for training OpenAI's model, so when entering personal information (customer name, address, etc.) into ChatGPT, the consent of the person identified by the personal information is required. must be obtained. In principle, entering personal information into other tools is prohibited. Please contact the corporate department for details.

(5) Confidential information disclosed by another company under a confidentiality obligation

The act of inputting confidential information obtained by concluding a non-disclosure agreement (NDA) etc. with another company into a generated AI provided by an external company is a risk of transmitting confidential information to a "third party" called the generated AI provider. This would be a "disclosure" and may violate the NDA.

Therefore, please do not enter such confidential information.

(6) Confidential information of own organization

The act of inputting confidential information within a company (know-how, etc.) into a generation AI does not violate any laws or regulations, but depending on the processing content of the generation AI and the content of the terms and conditions, the confidential information may no longer be protected by law or may be subject to patent protection. Please do not enter any information as there is a risk that you may not be able to apply.

6 Points to note when using the product

(1) The content of the product may contain falsehoods

The principle of a large-scale language model (LLM) is to create plausible sentences by outputting the word that has the highest probability of being used next to a certain word. The written content may contain falsehoods.

Please be aware of these limitations of generated AI, do not blindly trust the contents of the generated AI, and be sure to check the basis and support yourself.

(2) The act of using the product may infringe on someone's existing rights.

　　① Copyright infringement

　　If the product output using generated AI is the same or similar to an existing copyrighted work, the act of using the product (copying, distribution, etc.) may fall under copyright infringement. there is.

　　Therefore, please observe the following notes.

　　- Do not use specialized AI that has learned only the works of a specific author or writer.

　　- Please do not enter existing works, authors' names, or works' names in the prompt.

　　- In particular, when "using" a product (distribution, publication, etc.), it is necessary to investigate whether the product is similar to existing works and to ensure that the use of the product is subject to rights restriction regulations (such as Article 30, Paragraph 1 of the Copyright Act and similar regulations). Please consider whether it falls under Article 30-3, etc.).

　　② Trademark/design rights infringement

　　The act of using images generated using image generation AI or catchphrases generated using text generation AI for product logos, advertising, etc. is a registered trademark or registered design that is owned by another party. There is a possibility of copyright infringement, so in addition to investigating whether the product is similar to existing copyrighted works, please also conduct a search for registered trademarks and registered designs.

　　③ False personal information, defamation, etc.

　　ChatGPT and others are known to potentially generate false information about individuals. Acts of generating, using, and providing false personal information may violate the Personal Information Protection Act (violation of Articles 19 and 20 of the Act), as well as defamation and damage to credibility, so such acts are prohibited. Please do not do this.

　　(3) Copyright may not apply to the product.

　　If a product is not copyrighted, the product can basically be imitated by a third party, so individuals and organizations that require protection of their rights as their own creations. It's a big problem for.

　　Regarding this issue, the conclusion will differ depending on whether or not there is a "creative contribution" by humans to the creative activities using generative AI, so avoid using the product as is, and add and modify it as much as possible. Please be so.

　　(4) The product may not be commercially available.

　　When using products generated by generation AI in business, the question is whether the products can be used commercially.

This issue varies depending on the terms of use and contract conditions of the generation AI you use, so please use commercially available tools for each tool.

　　(5) Be aware of policy restrictions on generative AI

　　In addition to the risks explained so far (mainly legal restrictions) for generated AI, unique restrictions may be set in accordance with service policies.

When using ChatGPT, please note the following points.

　　　　Usage Policies (https://openai.com/policies/usage-policies) includes "Adult content, adult industries, and dating apps" and "Engaging in the unauthorized practice of law," There are specific prohibitions such as "or offering tailored legal advice without a qualified person reviewing the information." It is determined.

　　　　Additionally, when creating and serving content for consumers, such as those in the medical, financial, and legal industries, news generation, and news summaries, we also provide disclaimers that inform users of the use of AI and its potential limitations. The policy also states that the information must be provided to

　　　　Furthermore, related policies stipulate that when publishing content generated using OpenAI's services such as ChatGPT, it must be clearly stated that the content is generated using AI.

　　　　When using other tools, please understand and comply with the policies within the tools.