Every day brings new news about AI agents going “rogue.” AI models are increasingly acting like the bad guys, including the Hugging Face breach, hacking gym websites, and creating their own fake profiles for social engineering intrusions.
That’s why AI labs that create hacking models are expanding their cyber protection offerings. This week, OpenAI announced an expansion of Daybreak, the cyber defense service it launched earlier this year, not long after Anthropic released its cyber-focused model Mythos.
Daybreak is a service that bundles access to models, tools, and workflows for defenders. This expansion includes access to a brand new cyber-focused model designed for defense work.
OpenAI announced Monday that Daybreak will consist of two tiers: Blue and Red. Both of these tiers grant authorized customers access to OpenAI’s limited-access frontier cyber models. The Frontier model (the most advanced model available) has been the subject of debate. The Trump administration had previously sought cooperation with AI companies to deploy such models, citing safety concerns. Previously, OpenAI introduced significant guardrails for the use of these models, limiting what customers could do with them.
Although Blue appears to be the more basic of the two, it offers a variety of cyber services such as incident response, malware analysis, and patch validation. OpenAI calls Blue the “recommended starting point for most defenders,” suggesting it is sufficient for most enterprises.
Red, on the other hand, offers a broader and potentially more dangerous toolkit. The company offers users “purpose-trained cybersecurity models” designed to perform security testing and vulnerability research.
Red also comes with a new model GPT‑5.6‑Cyber available only in that tier. 5.6-Cyber is built on GPT-5.6 Sol, and the company says it offers enhanced capabilities for certain specialized cybersecurity tasks.
At this time, GPT‑5.6‑Cyber is only available to “trusted customer partners,” including Accenture, IBM, Crowdstrike, Cloudflare, and more.
While the threat posed by AI agents is rapidly increasing, critics also point out that AI agents serve as a marketing opportunity for AI laboratories. OpenAI is certainly marketing the upgraded Daybreak as such.
“The world of cybersecurity is rapidly changing, and threat actors will increasingly leverage AI to carry out cyberattacks at unprecedented speed and scale, including in fully autonomous ways,” the company said in a blog post. “As these capabilities expand, the window of time defenders have to prepare is shrinking.”
At the same time, companies are still interested in purchasing protection from the AI labs that know security risks best, because they know them first-hand.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
