OpenAI Restricts GPT-5.6 Sol Preview Amid Cybersecurity Misuse Concerns
OpenAI has launched a limited preview of its GPT-5.6 model family—Sol, Terra, and Luna—with access initially restricted to a small group of trusted partners through the API and Codex. The company described Sol as its most advanced cybersecurity-focused model and said the rollout follows consultations with the U.S. government while broader national-security risk assessment frameworks for cyber-capable AI are developed. OpenAI said wider availability across ChatGPT, Codex, and API offerings is planned in the coming weeks.
OpenAI said GPT-5.6 Sol improves vulnerability discovery, patch development, and exploit-related research, and internal as well as third-party testing indicated it can identify security flaws, generate credible memory-safety leads, and in some cases uncover previously unknown vulnerabilities. At the same time, the company said the model is not yet capable of reliably carrying out autonomous end-to-end attacks against hardened targets, and it has added layered safeguards including refusal training, output screening, real-time classifiers, secondary review models, account-level evaluations, misuse monitoring, and large-scale automated red-teaming to limit abuse of sensitive cyber capabilities.
Get ahead of threats like this
Mallory correlates global threat intelligence with your attack surface — know if you’re exposed before adversaries strike.
How this story unfolded
3 events from the most recent confirmed update back to the earliest known activity.
OpenAI reports GPT-5.6 still falls short of autonomous full attack chains
OpenAI stated that although GPT-5.6 Sol can generate credible vulnerability leads and assist with exploitation workflows, it is not yet capable of carrying out autonomous end-to-end attacks against hardened or well-defended targets. Third-party and internal testing cited in the reports found the model stronger at partial offensive tasks and defensive research than at weaponizing vulnerabilities in real-world attacks.
OpenAI discloses GPT-5.6 Sol cyber capabilities and safeguards
OpenAI said GPT-5.6 Sol is its most advanced cybersecurity model, with stronger performance in vulnerability discovery, exploit development, and patch-related defensive tasks than prior models. The company also disclosed layered safeguards including refusal training, output screening, misuse monitoring, real-time classifiers, secondary review models, account-level evaluations, and large-scale automated red-teaming.
OpenAI begins limited preview rollout of GPT-5.6 models
OpenAI announced a limited preview rollout of its GPT-5.6 lineup—Sol, Terra, and Luna—to a small group of trusted partners via the API and Codex. The restricted release followed consultations with the U.S. government and was framed as a temporary measure while broader AI risk assessment frameworks are developed.
Related entities
Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.
Sources
7 references tracked. Mallory keeps watching after this page renders.
GPT-5.6 Sol Debuts With Enhanced Cyber Protections, Limited to Trusted Partners - CySecurity News - Latest Information Security and Hacking Incidents
cysecurity.news
Open sourceOpenAI reporte GPT-5.6 Sol : l'ingérence de Washington bouscule à ...
zdnet.fr
Open sourceOpenAI Unveils GPT-5.6 Sol as Its Most Advanced Cybersecurity AI - SecurityWeek
securityweek.com
Open sourceGPT-5.6 gets better at cybersecurity - Help Net Security
helpnetsecurity.com
Open sourceOpenAI открыла ограниченный доступ к GPT-5.6 - Хакер
xakep.ru
Open sourceOpenAI Previews GPT-5.6 Sol With Restricted Access and Stronger Cyber Safeguards
thehackernews.com
Open sourceteiss - News - OpenAI defers public rollout of GPT‑5.6 as US seeks early access to frontier AI models
teiss.co.uk
Open sourceSee the full picture, correlated to your attack surface.
Map indicators from this story to your assets and identify affected systems in minutes.
Every observed campaign, victim, and pivot linked to actors named in this story.
Malware, exploits, and IOCs connected to the activity described here.
YARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Get matching new stories delivered to your team as they break — not the next morning.
Ask questions about this story and take action on the answers.
Related stories
Other stories Mallory is tracking that overlap on entities, sources, or topic labels.
OpenAI Releases Codex Security Agent and GPT-5.4 With Expanded Safety Controls
OpenAI announced *Codex Security* (formerly **Aardvark**), an application security agent designed to autonomously **discover, validate, and remediate** vulnerabilities in enterprise and open-source codebases. The product builds a project-specific threat model (mapping trust boundaries and exposure points), then prioritizes issues by likely real-world impact and attempts to **confirm findings by running proof-of-concept exploits in sandboxed environments** before generating contextual patches intended to minimize regressions. OpenAI reported beta metrics including an **84% reduction in alert noise**, a **90% decrease in over-reported severity**, and a **>50% drop in false positives**, and said the system scanned **1.2 million commits** from external repositories over a 30-day period; the tool is rolling out as a research preview via the Codex web interface to ChatGPT Pro, Enterprise, Business, and Edu customers. Separately, OpenAI released the **GPT-5.4** model across ChatGPT, Codex, and the API (`gpt-5.4` and `gpt-5.4-pro`), positioning it as a flagship model with improved reasoning, coding, and agent workflows, plus native computer-use capabilities. OpenAI stated it reduced hallucinations and errors (claiming user-flagged factual errors were **33% less likely** than GPT-5.2 on a de-identified prompt set) and said it **upgraded safety protections** while keeping the same high cyber-risk classification used for GPT-5.3-Codex, indicating continued emphasis on security controls as model capabilities expand.
Jun 29, 2026
OpenAI GPT-5.2-Codex Release and AI-Driven Vulnerability Detection
OpenAI has released GPT-5.2-Codex, a new AI model designed to enhance agentic coding and cybersecurity tasks, with notable improvements in vulnerability detection and software engineering workflows. The model demonstrates superior performance on benchmarks such as SWE-Bench Pro and Terminal-Bench 2.0, and excels in professional Capture-the-Flag challenges, supporting advanced tasks like fuzzing, attack surface analysis, and test environment setup. OpenAI has implemented stronger safeguards to mitigate dual-use risks and is offering the model to paid ChatGPT Codex users, with an invite-only pilot for vetted cybersecurity professionals focused on defensive applications. The model's capabilities have already contributed to the discovery of several critical vulnerabilities in React Server Components, including CVE-2025-55182 (RCE), CVE-2025-55183 (source code exposure), and CVE-2025-67779 (DoS), prompting urgent patching recommendations. Industry research highlights both the promise and risks of AI-generated code, with studies showing that machine-written pull requests contain significantly more bugs and security vulnerabilities than those authored solely by humans. Issues such as improper password handling, insecure object references, and cross-site scripting are notably more prevalent in AI-generated code, raising concerns about the security implications of rapid AI-driven development. These findings underscore the importance of robust review processes and the need for continued vigilance as AI tools become more deeply integrated into software engineering and cybersecurity operations.
Jun 29, 2026
GPT-5.5 Matches or Surpasses Mythos in Offensive Cybersecurity Tests
The UK’s AI Security Institute found that OpenAI’s **GPT-5.5** performed at or above the level of Anthropic’s **Mythos Preview** in offensive cybersecurity evaluations, underscoring a sharp rise in frontier-model cyber capability. In an AISI benchmark built around **95 Capture The Flag challenges**, GPT-5.5 reportedly led overall on advanced tasks with a **71.4%** score, ahead of Mythos Preview at **68.6%** and GPT-5.4 at **52.4%**. The tested scenarios included realistic security work such as firmware analysis without source code, exploitation of memory-corruption bugs including `use-after-free`, cryptographic bypass, attacks on weak random-number generators, and analysis of obfuscated malware. The results also showed limits: Mythos Preview reportedly remained stronger in longer, multi-step intrusion simulations, including a **32-step** exercise known as **“The Last Ones,”** and neither model completed a simulated industrial-system attack. Researchers and commentators said the findings point to rapid gains driven by greater autonomy, stronger reasoning, and improved coding ability, increasing concern over near-term misuse. Reports differed on deployment details, with some noting restrictions on GPT-5.5’s full public availability while others described the model as generally available, but the evaluations broadly agreed that leading AI systems are now highly capable at vulnerability discovery and offensive security tasks.
Jun 29, 2026