Get the weekly research letter
Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean. No account.
The policy prohibits using OpenAI services for weapons development, procurement, or use, including both conventional weapons and chemical, biological, radiological, nuclear, or explosive (CBRNE) materials.
This analysis describes what OpenAI's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology
This provision establishes a categorical prohibition on using OpenAI services for weapons-related purposes including CBRNE, which is operationally significant for defense contractors, research institutions, and security firms evaluating the scope of permissible use of OpenAI tools.
Interpretive note: The prohibition on 'conventional weapons' is not further defined, and the national security exception pathway is referenced but not procedurally described, creating uncertainty about the scope of permissible security-related use cases.
This provision establishes that OpenAI services may not be used for any weapons development, procurement, or operational use, including both conventional and CBRNE weapons categories.
Cross-platform context
See how other platforms handle Prohibition on Weapons Development and CBRNE and similar clauses.
Compare across platforms →Monitoring
OpenAI has changed this document before.
Receive same-day alerts, structured change summaries, and monitoring for up to 25 platforms.
"Protect people. Everyone has a right to safety and security. So you cannot use our services for: threats, intimidation, harassment, or defamation suicide, self-harm, or disordered eating promotion or facilitation sexual violence or non-consensual intimate content terrorism or violence, including hate-based violence weapons development, procurement, or use, including conventional weapons or CBRNE illicit activities, goods, or services destruction, compromise, or breach of another's system or property, including malicious or abusive cyber activity or attempts to infringe on intellectual property rights of othersExcerpt from OpenAI's Usage Policies
(1) REGULATORY LANDSCAPE: This provision engages export control frameworks including the Export Administration Regulations and the International Traffic in Arms Regulations in the US, as well as international arms control treaties governing CBRNE materials. The EU AI Act classifies certain AI applications in defense and national security contexts as high-risk or prohibited. Relevant enforcement authorities include the Department of Commerce Bureau of Industry and Security, the Department of State Directorate of Defense Trade Controls, and the DOJ for domestic law enforcement purposes. (2) GOVERNANCE EXPOSURE: High. Defense contractors, dual-use technology firms, and security research organizations must evaluate whether their intended use cases fall within this prohibition. The prohibition covers 'conventional weapons' without further definition, which may create uncertainty for legitimate security research, law enforcement training, or historical and academic analysis involving weapons-related content. (3) JURISDICTION FLAGS: US defense and national security contexts are most directly implicated given export control frameworks. International operators must evaluate compliance with applicable arms control and dual-use technology regulations in their jurisdictions. The policy separately prohibits 'national security or intelligence purposes without our review and approval,' which creates a distinct pathway for approved national security use but does not specify approval procedures. (4) CONTRACT AND VENDOR IMPLICATIONS: Defense and security sector organizations considering OpenAI API integration should assess this provision against their specific use cases and obtain clarity from OpenAI regarding the national security review and approval pathway referenced in the policy. The policy does not disclose the criteria or process for national security approval. (5) COMPLIANCE CONSIDERATIONS: Legal and export control teams at organizations working in defense, dual-use technology, or security research should conduct use-case reviews against this provision. The carve-out for national security purposes 'with our review and approval' suggests an exception pathway exists but is procedurally undefined in the policy text.
This provision establishes a categorical prohibition on using OpenAI services for weapons-related purposes including CBRNE, which is operationally significant for defense contractors, research institutions, and security firms evaluating the scope of permissible use of OpenAI tools.
This provision establishes that OpenAI services may not be used for any weapons development, procurement, or operational use, including both conventional and CBRNE weapons categories.
No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by OpenAI.