With the current launch of GPT 5.2, OpenAI up to date different associated fashions, together with its well-liked coding mannequin Codex, bringing extra agentic use circumstances to its fold.
GPT-5.2-Codex, which OpenAI referred to as in a weblog put up “essentially the most superior agentic coding mannequin but for complicated, real-world software program engineer,” has been optimized for long-horizon work with brokers and could have stronger cybersecurity capabilities.
The mannequin is an offshoot of GPT-5.2, optimized for agentic constructing.
“GPT‑5.2-Codex represents a step ahead in how superior AI can help real-world software program engineering and specialised domains like cybersecurity—serving to builders and defenders sort out complicated, long-horizon work, and strengthening the instruments out there for accountable safety analysis,” the corporate mentioned in its weblog put up.
Enterprises can entry the brand new Codex mannequin “in all Codex surfaces for paid ChatGPT customers, and dealing in the direction of safely enabling entry to GPT‑5.2-Codex for API customers within the coming weeks.” The corporate can be piloting a program with invite-only trusted customers to entry “extra permissive fashions for vetted professionals and organizations” for defensive cybersecurity work to find out a stability between accessibility and security.
Advances in cybersecurity with fashions
OpenAI calls GPT-5.2-Codex its strongest cybersecurity mannequin but. Nonetheless, as its capabilities develop, the corporate mentioned it must design a deployment strategy that accounts for future development and helps defensive cybersecurity.
“As our fashions proceed to advance alongside the intelligence frontier, we’ve noticed that these enhancements additionally translate to functionality jumps in specialised domains comparable to cybersecurity,” the corporate mentioned.
OpenAI mentioned in its system card that it examined the mannequin on three benchmarks: Seize-the-Flag (CTF) evals, CVE-Bench and Cyber Vary.
GPT-5.2-Codex grew to become the corporate’s strongest-performing mannequin in CTF evals, which they attributed to compaction, or “the flexibility for the mannequin to work coherently throughout a number of context home windows.”
The mannequin scored 87% in CVE-Bench, outperforming different fashions, with GPT-5.1-Codex-Max coming in an in depth second. This improve could be useful for duties involving operating instructions round vulnerability discovery and making an attempt instruments “with an virtually brute-force strategy.”
Within the long-form Cyber Vary check, the mannequin had a mixed go fee of 72.7%. GPT-5.1-Codex-Max scored 81.8%.
Cybersecurity deployment challenge
OpenAI mentioned some customers of its GPT-5.1-Codex-Max, which launched in November, uncovered a supply code publicity vulnerability in React and subsequently reported it. In keeping with OpenAI, Andrew MacPherson, a safety researcher at Privy, used GPT-5.1-Codex-Max to evaluate how effectively the mannequin might help real-world vulnerability analysis. The mannequin as an alternative surfaced surprising habits.
With enhancements in cybersecurity capabilities for GPT-5.2-Codex and probably for fashions that come after it, OpenAI mentioned it must stability the deployment of frontier fashions with the mandatory instruments for defensive cybersecurity. Whereas GPT-5.2-Codex “doesn’t attain a excessive degree of cyber functionality underneath our Preparedness Framework,” the corporate plans to carry chosen customers to check safety capabilities. (OpenAI’s Preparedness Framework to measure and observe potential harms from AI to people)
“Safety groups can run into restrictions when making an attempt to emulate risk actors, analyze malware to help remediation, or stress check vital infrastructure. We’re growing a trusted entry pilot to take away that friction for qualifying customers and organizations and allow trusted defenders to make use of frontier AI cyber capabilities to speed up cyberdefense.” OpenAI mentioned.
Agentic frontiers
GPT-5.2 already acquired reward from customers for its use in enterprise duties and workflows. With the Codex model, a few of these capabilities might switch, particularly as enterprises plan to make use of the mannequin to code their brokers.
The corporate mentioned the mannequin improves long-horizon work by compaction, providing sturdy efficiency on in depth code adjustments. It additionally options improved efficiency on Home windows.
In benchmark testing, GPT-5.2-Codex carried out the perfect on accuracy in comparison with its earlier variations.
"With these enhancements, Codex is extra succesful at working in massive repositories over prolonged periods with full context intact. It could actually extra reliably full complicated duties like massive refactors, code migrations, and have builds — persevering with to iterate with out shedding observe, even when plans change or makes an attempt fail," OpenAI mentioned.
Because it launched in previews in Might, Codex has helped usher in acceptance of agentic and vibe coding within the enterprise AI builder house. Together with Windsurf, Cursor, Claude Code and the various coding brokers from Google, the platform moved LLMs from easy code completion to producing and beginning asynchronous coding tasks for customers.
[/gpt3]
