New Delhi: OpenAI has launched a new cybersecurity initiative, called Daybreak, which intends to enable developers, enterprises and government-related
organisations to find and fix software vulnerabilities sooner with the help of AI. The programme is designed around Codex Security and industry-specific AI models based on GPT-5.5, as OpenAI continues to push AI as the foundation of enterprise security operations – not just a productivity tool.
The launch follows an era in which cyber threats are growing more complex and speedy. Rather than being a mere tool for passive assistance, OpenAI claims Daybreak aims to actively aid in threat modelling, vulnerability validation, remediation testing, and secure software development. On X, OpenAI CEO Sam Altman called Daybreak its “attempt to speed up cyber defence and continually secure software”.
OpenAI wants AI embedded inside cyber defence operations
Introducing Daybreak: frontier AI for cyber defenders.
Daybreak brings together the most capable OpenAI models, Codex, and our security partners to accelerate cyber defense and continuously secure software.
A step toward a future where security teams can move at the speed… pic.twitter.com/AGfXhmJb5E
— OpenAI (@OpenAI) May 11, 2026
The project’s core is the company’s application security tooling, Codex Security, which is an extended version of OpenAI’s tooling. It can be used to scan company code bases, find possible attack paths, test vulnerabilities within isolated environments and suggest patches for human review.
Daybreak is OpenAI’s attempt to expand the scope of AI from a developer-friendly coding assistant to an integrated, enterprise-level defence system alongside the network’s security teams. The aim of the company is to enable organisations to identify flaws earlier in the software development process, rather than responding after vulnerabilities are exploited.
Secure code review, dependency risk analysis, remediation guidance and support for detection engineering workflows in the same location as code in Codex Security are also part of the initiative.
Trusted access system separates general and advanced cyber capabilities
OpenAI is launching Daybreak, our effort to accelerate cyber defense and continuously secure software.
AI is already good and about to get super good at cybersecurity; we’d like to start working with as many companies as possible now to help them continuously secure themselves.
— Sam Altman (@sama) May 11, 2026
In addition to this, OpenAI has launched a new OpenAI Trusted Access for Cyber framework to support the rollout. In such a system, the use of AI in the cyberspace will be determined by the sensitivity of the tasks carried out.
For users who don’t need the power and capabilities of GPT-5.5, the standard GPT-5 models will continue to be accessible. But GPT-5.5 with Trusted Access is designed for trusted security professionals to use in the analysis of malware, triage of vulnerabilities, validation of patches and secure code review. OpenAI says these users will receive stronger cyber assistance with fewer unnecessary refusals.
The company is also conducting tests for GPT-5.5-Cyber, a small preview version designed for specific processes like authorised penetration testing, controlled red teaming and advanced validation exercises.
Meanwhile, OpenAI notes that there are still restrictions in place to prevent misuse that includes credential theft, malware deployment, stealth attacks and unauthorised use.
OpenAI partners with major cybersecurity firms
Daybreak will receive support from several major cybersecurity and infrastructure companies. Cloudflare, Cisco, CrowdStrike, Palo Alto Networks, Oracle, Zscaler and Fortinet are among OpenAI’s partners.
The wide range of partner ecosystems indicates that OpenAI is looking for Daybreak to be used throughout the entire cybersecurity lifecycle, from vulnerability discovery through patch validation to monitoring and software supply chain protection.
The programme is still closed for now. Organisations are encouraged to seek vulnerability scans or reach out directly to OpenAI’s enterprise teams; broader rollouts will take place over time via industry and government partnerships.
OpenAI balancing cyber defence and misuse risks
The slow deployment further underscores the rising worries about dual-use AI systems. If left unchecked, the same AI models that can assist defenders in securing infrastructure can be leveraged to carry out offensive cyber activities.
To respond to those concerns, OpenAI adds more robust verification systems, limited permissions, controls on the account level, monitoring and human oversight features to Daybreak.
Daybreak is, for OpenAI, more than just a cybersecurity product launch. The move reflects the firm’s broader strategy to make its AI platforms into regulated enterprise platforms that can run within a sensitive digital ecosystem.
