OpenAI GPT-6 Astra Reaches 100% Exploit Generation Accuracy on ExploitBench
OpenAI has introduced GPT-6 Astra, describing it as the most intelligent and human-aligned model to date. The system can operate computers and browsers, write code, solve scientific problems, and generate exploits with exceptional precision, prompting the company to withhold an unrestricted release.
On the ExploitBench benchmark, which evaluates an AI's ability to convert known vulnerabilities into functional exploits, Astra achieved a perfect score of 100%. This marks a substantial improvement over the previous GPT-5.6 Sol, which reached only 78.5%.
In separate evaluations using newly disclosed vulnerabilities, the model successfully executed arbitrary code in 39% of attempts, compared to 11.5% for its predecessor. During these tests, Astra also identified two previously unknown zero-day vulnerabilities, which OpenAI has shared with the affected software vendors.
Expert testing without public safeguards showed that Astra can exploit unknown vulnerabilities to achieve code execution in hardened browsers and develop privilege-escalation exploits for modern operating systems. Consequently, the publicly available version is limited to secure code analysis and patch preparation, while requests to create PoC exploits are automatically rejected.
Advanced access for vulnerability assessment, malware analysis, and detection tool development will be distributed through the Daybreak program. OpenAI has also strengthened jailbreak defenses and implemented continuous activity monitoring that may interrupt even legitimate cybersecurity tasks and require user confirmation.
The model is currently being rolled out to selected organizations. It will later become available to ChatGPT Plus, Pro, Business, and Enterprise subscribers, as well as through the API, Microsoft Azure, and AWS Bedrock. In parallel, OpenAI has committed $1 billion to supply advanced AI tools to energy providers, banks, water utilities, government agencies, and open-source developers.
Related articles
How to Interact with AI Models Without Exposing Sensitive Data
The article provides practical guidance on minimizing data leakage risks when using popular AI chatbots such as ChatGPT, Gemini, Claude and GigaChat. It explains that conversations are routinely scanned by automated filters and may be reviewed by human moderators or shared with law enforcement upon request. Key recommendations include disabling model training on user data, replacing sensitive values with placeholders, regularly deleting chat histories and verifying downloaded models for malicious injections. The guide also demonstrates local deployment using Ollama and secure API integration through the ChatBox client with Cloud.ru’s Evolution Foundation Models service. Local execution in Docker containers is presented as the most private option, although it requires significant computational resources. The author stresses that even after disabling training, data may still reach moderators and that users remain responsible for their own information.
Raft Develops Multilabel Guardrail Classifier Detecting 15 Risk Categories with 3x Speed and Cost Gains
Raft has released details on a custom multilabel guardrail classifier designed to scan both incoming prompts and model outputs for 15 distinct risk categories in real time. The system handles Russian and English text while maintaining independent thresholds for each category to balance false positives against critical misses. By switching from a PyTorch baseline to TensorRT inference on NVIDIA RTX 3090 hardware, the team achieved a 2.95x reduction in single-request latency and lowered inference cost to $0.062 per million requests. The architecture uses per-category expert query tokens plus a lightweight interaction transformer to capture correlations such as those between armament and violent content. Training relied on asymmetric loss functions and post-epoch per-class threshold tuning rather than standard binary cross-entropy. Benchmarking against nine open guardrail and toxicity models showed superior macro-F1 on rare but high-impact categories while remaining an order of magnitude cheaper than external LLM judges.
Zhou Hongyi Warns AI Tools Are Industrializing Vulnerability Discovery
At the Fourth Cyberspace Security Forum in Tianjin, 360 founder Zhou Hongyi stated that vulnerability mining is shifting from artisanal workshops to automated production lines, compressing discovery cycles from months or years down to hours. AI tools such as Mythos are standardizing and automating the process, enabling attackers to replicate elite hacker expertise at scale through distilled models and agent swarms. 360's own Tulongfeng platform has already discovered over 10,000 vulnerabilities since its June release, including long-hidden high-risk flaws in Windows, Office, OpenClaw, Flowise, and Codex. The emergence of multi-agent systems introduces new attack surfaces because compromised agents can autonomously collaborate and move laterally faster than human operators. Zhou described this as the "second one-way transparency," where offensive tradecraft becomes copy-pasteable via prompts and toolchains. Defenders are advised to adopt "model-versus-model" strategies, automate vulnerability intelligence workflows with SOAR, enforce strict agent permission audits, and integrate AI into their own code review and detection engineering processes.
Anthropic Fable 5.1 System Prompt Fully Leaked Hours After Launch Exposing 275000 Characters of Rules
Anthropic released its flagship Fable 5.1 model alongside Mythos 5.1 on September 2, achieving strong benchmark scores including 90 percent on ARC-AGI-2. Within hours, researcher Pliny the Liberator published the complete 275000-character system prompt on GitHub, far exceeding the company's official 27000-word disclosure. The leaked document details 46 built-in tools, strict copyright restrictions, memory classification boundaries, and behavioral constraints that function as an internal employee handbook. The incident highlights that model weights remain the true core while prompt-based guardrails create an attack surface once mapped. It also reveals privacy rules that permanently exclude storage of minor identities, criminal records, and self-harm indicators even when users disclose them. The leak underscores the growing gap between vendor transparency claims and actual runtime instructions governing frontier AI systems.