March 27, 2026 Results of the public comment solicitation on the "Guidelines on Technical Measures to Ensure the Security of AI" (draft) and publication of the guidelines

　The Ministry of Internal Affairs and Communications solicited public comments on the "Guidelines on Technical Measures to Ensure the Security of AI" (draft) from Friday, December 26, 2025, to Thursday, January 29, 2026. Based on the results of this public comment solicitation, we have formulated the "Guidelines on Technical Measures to Ensure the Security of AI" and hereby publish them.

*The published materials are in Japanese only.

1 Overview

　Since September 2025, the Ministry of Internal Affairs and Communications has convened the AI Security Subcommittee under the Cyber Security Task Force to examine technical measures for ensuring AI security against threats such as causing AI to produce malicious outputs or causing systems equipped with AI to shut down. We solicited the public comments on the "Guidelines on Technical Measures to Ensure the Security of AI" (draft) drafted based on the subcommittee’s findings, from Friday, December 26, 2025, to Thursday, January 29, 2026, and received 52 submissions.

　Based on the public comments received, we have finalized the "Guidelines on Technical Measures to Ensure the Security of AI" and publish along with the submitted the public comments and the Ministry of Internal Affairs and Communications’ responses to the public comments.

2 How to Obtain the Documents

　The published materials are available in the "Public Comments" section on the e-Gov website: https://www.e-gov.go.jp