Chapter 3 describes usage rules. If you use articles to generate AI, be sure to read this chapter and follow the rules for safe use.
terms of service
When utilizing article-generating AI, it is important to maintain a safe usage environment, as well as for employees themselves to understand the risks and follow the rules when using it. First, be sure to observe the following points. Shared use of article generation AI within institutions that ensure security (see Chapter 2) The Tokyo Metropolitan Government Cybersecurity Policy prohibits the unauthorized use of generation AI for business purposes, and the use of generation AI on personal devices is prohibited. Doing business is also prohibited and violations may result in disciplinary action. In addition, we ask all employees to start using the service after performing the following operations. Please be sure to apply for e-learning from the application form before using it, and use it correctly.
Rules for employees to follow
Rule 1 Do not enter sensitive words such as personal information Rule 2 From the viewpoint of copyright protection, please be careful and confirm the following points. Do not enter immediate information regarding the creation of articles similar to existing works. Be careful when sending or publishing responses. Be sure to confirm the basis and evidence of your answer
Rule 1 Do not enter confidential information such as personal information
Text generation AI is an external service, and security measures depend on the service provider. Entering public information leaves the risk of information leakage in the worst case.
* Confidentiality A cannot be entered (if the usage environment can ensure security (see Chapter 2), confidentiality B information can be entered)
From the viewpoint of copyright*1 protection, it is necessary to pay sufficient attention to and confirm the following points.
・Existing (1) Sentences similar to copyrighted works Create sentences similar to existing copyrighted works *Please do not enter the prompt leading to 2.・When distributing or posting a reply, please make sure that it is not similar to existing copyrighted works. The simple act of entering someone else's existing work, the author's name, or the title of the work does not necessarily constitute immediate copyright infringement. However, if the data generated is identical or similar to prompted data or an existing work, use of such products may constitute copyright infringement of that work. Especially if you're distributing or posting a product, find out if it resembles that crop. *1 Including copyrights, patent rights, and individual contract rights. *2 Enter "Please generate a sentence similar to the work XX of XX artist." In some cases, we may request copyrighted material and generate text similar to that copyrighted material.
Please check with your own eyes the grounds and grounds for the answers generated by the sentence generation AI.
The answers generated by the sentence generation AI are natural expressions and phrases, so I think they are correct. However, it may not reflect the latest information, or may reflect biased values, unconscious bias*, etc., and is not necessarily "accurate". * Distortions and biases in the way you see and perceive things that you are not aware of
If the AI response sentence generation is directly used externally, specify that fact
After confirming the content, if you want to directly use the response of the text generation AI externally such as translation or summary, you can check whether the generated text is AI generated by declaring "Created by text generation AI" . You can tell the reader if it was made by humans or not.