Daybreak models are now available on AWS
OpenAI has made its Daybreak cybersecurity models available on Amazon Web Services via Amazon Bedrock, enabling enterprises to integrate frontier AI into existing security workflows.
OpenAI announced that its Daybreak cybersecurity models are now accessible through Amazon Bedrock, expanding the company’s collaboration with AWS. The models, including Daybreak Blue and Daybreak Red, are designed to assist security teams in vulnerability research, detection engineering, and incident response. Enterprises can deploy these capabilities within their existing AWS environments, leveraging familiar security and governance frameworks. The integration aims to streamline the adoption of advanced AI tools for cybersecurity operations.
Daybreak Access, which includes both Daybreak Blue and Daybreak Red, requires enrollment and approval before use. Once granted, customers can access the models through the Amazon Bedrock console or the Responses API using the bedrock-mantle endpoint. This approach ensures that security teams can integrate frontier AI without disrupting established workflows. OpenAI and AWS emphasize the practicality of deploying these models in production environments.
The models support complex cybersecurity tasks such as exploit reproduction and mitigation development, addressing critical needs in vulnerability management. By making Daybreak available on AWS, OpenAI and Amazon provide a pathway for enterprises to adopt specialized AI capabilities securely. The collaboration underscores the growing role of AI in enhancing cybersecurity operations across industries.
Eligible customers can now use Daybreak within their AWS environments, aligning with existing security, governance, and operational processes. OpenAI highlights that the integration simplifies procurement, access controls, and governance for enterprises. The move reflects a broader trend of embedding advanced AI tools into enterprise security infrastructures.