"Security Tips for AI Users" and "AI Security Brief" Released

The Information-technology Promotion Agency, Japan (IPA, President: Yu Saito) and the AI Safety Institute (AISI) have been discussing AI security with experts and industry organizations since 2025, aiming for the safe and secure utilization of AI. As a result of these activities, they have now publicly released "Security Tips for AI Users" (hereinafter, "Tips") for general users and "AI Security Brief" (hereinafter, "Brief") which summarizes the latest trends and information for AI developers and security personnel.

The "Tips" are provided in a presentation slide format that can be easily used for in-house training. They will be updated as needed to match the expansion of AI use cases and trends.

New issues of the "Brief" are planned to be published several times a year in the future.

■Background

In recent years, while the practical application of advanced AI technologies, including generative AI, has rapidly progressed, security issues such as AI misuse and attacks targeting vulnerabilities have become apparent. In the "Top 10 Information Security Threats 2026" released by IPA in January this year, "Cyber risks surrounding AI use" ranked 3rd among threats for "organizations." As AI utilization advances, there is a growing need to strengthen and promote security measures for both AI users and developers.

■Overview

"Security Tips for AI Users" is aimed at users who want to understand the basics of AI and security, summarizing the minimum security measures that should be implemented when using AI in an easy-to-understand manner. It is provided in a presentation slide format that can be used as is for in-house training and will be continuously updated to accommodate future expansions of AI use cases.

"AI Security Brief" is aimed at AI system developers and security personnel, introducing the latest trends and case studies related to AI security. It organizes the vast amount of information disseminated daily from an AI security perspective. Previously, it was distributed irregularly to experts and industry organizations, but in response to feedback from industry organizations who wanted more AI and security personnel to be aware of it, the latest issue and the three previously created issues have been released to the public. Given the current situation where there are many experts in AI or security, but not many who are proficient in both fields, we hope this will contribute to awareness.

Both materials will continue to be disseminated in the future. Please download and use them as appropriate.
For details, please refer to the IPA website.

■Inquiries regarding this matter

IPA Security Center, Planning Department

Email: [email protected]