Models · OpenAI
Cybersecurity is one of the clearest places where that progress can both meaningfully strengthen the broader ecosystem and introduce new
Compiled by KHAO Editorial — aggregated from 1 outlet. See llms.txt for citation guidance.
★ Tier-1 Source
To unlock the full defensive potential of these capabilities while reducing the risk of misuse, they are piloting Trusted Access for Cyber: an identity and trust-based framework designed to help ensure enhanced cyber capabilities are being placed in the right hands.
Key facts
- In addition, they are committing $10 million in API credits to accelerate cyber defense
- To further accelerate the use of their frontier models for defensive cybersecurity work, they are committing $10 million in API credits for teams through their Cybersecurity Grant Program
- GPT‑5.3‑Codex is their most cyber-capable frontier reasoning model to date
- Frontier models like GPT‑5.3‑Codex have been designed with mitigations like training the model to refuse clearly malicious requests like stealing credentials
Summary
Their approach to enhancing baseline safeguards for all users while piloting trusted access for defensive acceleration. GPT‑5.3‑Codex is their most cyber-capable frontier reasoning model to date. It is important the world adopts frontier cyber capabilities quickly to make software more secure and continue to raise the bar of security best practices. There will soon be many cyber-capable models with broad availability from different providers, including open-weight models, and they believe it is critical that OpenAI’s models strengthen defensive capabilities from the outset.