OpenAI launches GPT-5.6-Cyber, expands Daybreak initiative with 2 new tiers

OpenAI has unveiled its latest AI model, GPT-5.6-Cyber, and expanded its Daybreak cybersecurity initiative with two new tiers. This development comes amid a surge in cybersecurity incidents, with advanced AI models increasingly exhibiting unexpected, potentially malicious behavior. OpenAI's new cybersecurity-focused AI model and the expanded Daybreak initiative aim to provide trusted defenders with advanced cyber capabilities.

GPT-5.6-Cyber, backed by GPT-5.6 Sol, has been trained to excel in specialized cybersecurity tasks, including discovering zero-day vulnerabilities, developing exploit chains, and reducing refusals for high-risk tasks. Currently, this cyber model is accessible only to 'trusted customer partners,' which reportedly include companies like Crowdstrike, IBM, and Cloudflare.

Daybreak, OpenAI's cybersecurity initiative, now offers two access tiers: Blue and Red. Daybreak Blue, recommended for most defenders, provides access to frontier general-purpose models, including GPT-5.6 Sol. This tier supports various cybersecurity tasks such as vulnerability discovery, secure code review, malware analysis, incident response, and patch validation.

In contrast, Daybreak Red offers access to OpenAI's purpose-trained cybersecurity models for authorized vulnerability research, exploit validation, and security testing. Both tiers allow approved customers to access OpenAI's limited-access frontier cyber models.

OpenAI believes that empowering trusted defenders with frontier intelligence is the best way to deploy offensive AI capabilities at scale. However, the use of AI by bad actors has accelerated cyber attacks at an unprecedented pace, leaving defenders with little time to prepare and act. The company's efforts to address these challenges include the introduction of GPT-5.6-Cyber and the expansion of the Daybreak initiative with its two new tiers.