undermine its credibility. For example, a manuscript written by an external contributor using generative AI may contain inaccurate information. If such output is published under the organization’s name, the organization may find it difficult to avoid responsibility. Accordingly, it is important to share the organization’s generative AI policy in advance and obtain agreement from external partners, or to engage in discussion where there are differences of opinion regarding the policy. When commissioning specific tasks or deliverables—such as written content or design work—the organization may include a clause in the request or contract stating that “the organization’s AI policy must be complied with.” 5) Measures in the Event of an Incident ① If any issue arises in connection with the use of generative AI, it shall be reported immediately to the AI Officer. The report shall include, where relevant, information such as: – date and time of the incident; – name of the AI tool used; – the relevant output; – the specific problematic elements; – the prompt input used; – the nature and scope of any negative impact. ② The AI Officer shall promptly verify the facts and, where necessary, take emergency measures to prevent the further spread of harm. ③ The AI Officer shall convene the [Steering Committee] to develop the organization’s response. This process shall include a review of the cause of the issue, the scope of its impact, whether and to what extent the organization bears responsibility, relevant legal 100 101