OpenAI has launched GPT-5.1-Codex-Max, a specialized agentic coding model optimized for extended software engineering tasks.
This frontier model, now the default in Codex tools, handles multi-hour workflows through innovative compaction technology that manages millions of tokens across context windows.
Trained on real-world tasks such...
A critical authentication bypass flaw in Milvus Proxy (CVE-2025-64513) allows attackers to bypass all security checks.
Discovered by the HelixGuard Team on November 12, 2025, this issue affects popular versions of Milvus, an open-source vector database designed for AI workloads, including generative models.
Attackers...
The National Security Agency (NSA), along with the Cybersecurity and Infrastructure Security Agency (CISA) and several international partners, released a new guide on November 19, 2025, to help internet service providers (ISPs) and network defense teams combat cyber threats from bulletproof hosting (BPH) providers....
Microsoft has introduced a new feature in Teams that allows users to misreport messages identified as security threats, helping to reduce false positives in organizational communications.
This update, tied to Microsoft 365 Roadmap ID 501202, began rolling out in early September 2025 for targeted...
Large language models like GPT-4, Claude, and Gemini rely on safety guardrails to block harmful prompts, but a new technique called EchoGram can trick these defenses into approving dangerous inputs.
Developed by researchers at HiddenLayer in early 2025, EchoGram exploits weaknesses in how guardrails...
DoorDash, the popular food delivery platform, has disclosed a cybersecurity incident where an unauthorized third party accessed certain user information through a social engineering attack.
The company confirmed the breach in a public statement, emphasizing that no sensitive financial or identification data was compromised....