MARADMIN 602/24 - Guiding Principles for the Ethical Use of Artificial Intelligence by Communication Strategy and Operations

Scope

MARADMIN 602/24 sets the guiding principles for the ethical use of artificial intelligence (AI) by the Communication Strategy and Operations (COMMSTRAT) community. The principles balance AI efficiency against national security interests and public trust. Key aspects include safeguarding national security, transparency about AI-enhanced content, human oversight in information release, and acknowledgment of inherent AI limitations. The source page reads CANCELLED. MARADMIN 635/24 supersedes this message.

Audience

• Communication Strategy and Operations (COMMSTRAT) community. Apply the ethical principles to AI use.
• Public affairs officers and Marine Corps communicators. Mandatory adherence to the principles.
• Commanders relying on COMMSTRAT products. Verify products follow the principles.

Guiding principles

1. AI use aligns with the DOD Principles of Information. Released information is truthful and accurate. AI is barred from deception, manipulation, and misinformation.
2. AI use upholds the responsibility to protect information. Adherence to security, accuracy, privacy, and propriety standards. Only DoD-approved AI technologies are authorized.
3. Imagery and video are uniquely truthful mediums for conveying actual events. AI is barred from creating photo-realistic imagery, video, or news stories for public dissemination. AI use is limited to reviewing written products or streamlining authorized adjustments to visual information per the DoD Visual Information Style Guide. Products adjusted with AI annotate the adjustment in both caption and metadata (example, basic color correction done with AI).
4. Build and maintain proficiency in conventional content creation methods. Safeguard against over reliance on AI. Avoid atrophy of creative and technical skills.
5. AI use is authorized to automate processes and enhance understanding of the information environment.

Connection to higher policy

• DoD Interim Guidance on the Use of Generative AI, dated 2023-10-15.
• DON Guidance on the Use of GenAI and LLM, dated 2023-09-06.
• USMC MARADMIN communicating the availability of authorized NIPR and SIPR LLM capabilities, dated 2024-10-21.
• DoD Visual Information Style Guide.

Connection to Marine Corps doctrine

• MARADMIN 635/24. Guiding Principles for the Ethical Use of AI by COMMSTRAT. Supersedes this MARADMIN.
• MARADMIN 018/26. Enterprise GenAI Availability and Governance.
• MARADMIN 270/25. Available AI and Data Management Training Solutions.

Status

Released 2024-12-17, DTG R 131819Z DEC 24. Originator CMC CD Washington DC. Release authorized by SES April Langwell, Director of Communication, Communication Directorate. The source page reads CANCELLED. Use this entry for historical context only. The current ethics framework lives at MARADMIN 635/24 (12 days later) which supersedes this message with broader wording on the scope of permitted AI use.