Over the past few months, OpenAI has worked with customers and partners to align safety policies across the services that provide its models. It is also updating its Usage Policies to make those requirements clearer and easier for developers to follow. The context is expanding model capabilities: models such as GPT-6 Astra expand what developers and businesses can build, and OpenAI says it is working to keep safeguards effective and consistently applied across the services that offer its models.

Before deploying a model, OpenAI evaluates its capabilities and risks, trains it to follow safety requirements, and tests whether safeguards hold up when people deliberately try to bypass them. The company says its approach to safety and alignment combines safeguards designed to reinforce one another. After deployment, it monitors for potential abuse and investigates concerns. To mitigate cyber misuse, OpenAI trains models to refuse harmful requests and uses systems that look for suspicious patterns across interactions to decide how to respond. Trusted access programs add another layer of protection by setting eligibility requirements and conditions for access to the most powerful capabilities. Some safeguards are built into the model, while others depend on the systems and people around it, and both need to be effective whether people access models directly or through a partner.

Developers and businesses should be able to choose where they access models, with consistent safeguards across those services. OpenAI developed policies to align safeguards across providers and is working with each partner to ensure implementation meets or exceeds OpenAI‘s safety standards. These policies apply whether models are accessed directly through OpenAI, through partners that host them, or through reseller partners. Partners that provide access to OpenAI models must maintain safeguards consistent with OpenAI‘s, investigate potential violations, and take appropriate action when concerns are identified, including coordinating on enforcement so restrictions remain effective across services. OpenAI says putting this into practice means accounting for how different services work and how customers use them; it points to work with customers and partners on Private Safety Processing as an example of developing effective safeguards while giving customers greater choice and control over sensitive data.

OpenAI is updating its Usage Policies, effective November 20, 2026, to clarify the rules for safe model development and sensitive uses of AI. The updated policies require services that provide direct access to OpenAI models, such as model routers or AI gateways, to get OpenAI‘s review and approval before offering that access. OpenAI says it will keep explaining what is allowed, what is restricted, and how customers and partners can ask questions or resolve compliance concerns.

As models become more capable, OpenAI says it will keep improving how it detects, investigates, and responds to misuse. Customers can face misuse through their end users, compromised accounts, or leaked API keys, and OpenAI works with them to understand and address those problems. The company is introducing SafetyIDs, which customers can provide to help distinguish individual end users during an investigation without OpenAI learning those users’ identities. The goal is to respond to potential misuse more precisely while reducing disruption to others. Enforcement decisions are based on documented conduct and risk, applied consistently, and proportional to the circumstances. When access is restricted, OpenAI revisits the restriction if meaningful changes in conduct, commitments, or safeguards address the concerns behind the decision.

OpenAI frames the changes as a step toward broader access: more people should be able to benefit from AI, and clear rules plus consistent safeguards should make it easier for developers and businesses to build, compete, and choose the services that work best for them.