AI Briefing#cybersecurity
The latest AI and developer news about #cybersecurity, with the original source and a short summary.
The latest AI and developer news about #cybersecurity, with the original source and a short summary.
Unit 42 research reveals Chinese-speaking attackers autonomously compromised 10 organizations in semiconductor, defense, and civil society sectors using DeepSeek and Zhipu GLM agents in Aug-Sep 2026.
Cloudflare identifies six pillars for the Modern Security Model to counter autonomous AI agents that chain stolen credentials for lateral movement.
The moves mark the fourth strategic M&A transaction for Cohere this year, following acquisitions of Aleph Alpha and Reliant AI.
OpenAI's unreleased internal model scored 77.0% on ExploitGym, nearly double Anthropic's Claude Mythos 5.2 (39.2%) and significantly outperforming GPT-6 Cyber (52.6%).
The cybersecurity leader and AISLE co-founder joins to build the world's most trusted platform for sovereign AI.
Anthropic has released Claude Fable 5.2 and Mythos 5.2, lowering prices and improving output speed by up to 30%.
Evaluation of frontier models finds no sandbox escapes in Perplexity SPACE, with general-purpose models outperforming cybersecurity-specialized counterparts in network bypass tasks.
The model targets sustained agentic workflows and complex multi-step problems in software engineering and scientific research.
The multi-LLM agentic framework generated over six million lines of Lean code to mathematically prove the security of the actual Signal code in four weeks.
The startup combines GPT-6.1 Sol and GPT-5.6 with enterprise context to autonomously investigate and fix security risks, achieving 5x triage throughput for IT services clients.
The model features a 1M-token context window for complex software engineering and cybersecurity tasks, with a 50% price reduction ending November 30, 2026.
An autonomous AI penetration-testing agent accessed WorkOS sales and support data, exposing four security failures including weak credentials and lack of rate limiting.
It recorded a 0.07% out-of-scope attack rate while discovering 948 Linux kernel privilege escalation vulnerabilities.
OpenAI announces a partnership with Hacktron to find vulnerabilities in its systems, focusing on interactions with third-party services and expanding Hacktron's access to frontier models via Daybreak Red.
The model developed 948 Linux privilege escalation exploits and maintains 99.93% scope adherence without safety prompts.
The autonomous pipeline cost $1,213,430 to identify and patch the vulnerabilities, a fraction of the estimated $18.9M to $66.3M value based on past bug bounty rates.
GPT-5.6 Cyber is now available to approved customers in AWS GovCloud (US-East) via Daybreak Red, supporting authorized vulnerability research and exploit validation.
The risk of external input traversing multiple AI agents to result in code changes must be mitigated through execution context security.
Cisco combines network telemetry with OpenAI models, utilizing advances like Daybreak and GPT-6 Astra to correlate security signals at machine scale for faster vulnerability remediation.
The benchmark covers 91 CWEs with deterministic scoring, where the top model still fails 19% of tasks.
A hidden reference to Claude Mythos 5.5 appeared on Anthropic's site, described as a model for cybersecurity and biology research, with a launch possibly as early as next week.
Gemini 4 Carbon takes first place with an AA Index score of 59 and is prioritized for provision to cyber defense organizations.
Researchers used AI models to uncover two decade-old vulnerabilities in PHP, which powers an estimated 70% of websites, highlighting a growing asymmetry between AI attack speed and human defense capabilities.
The funding, led by Schwarz Group, aims to develop frontier models and secure cyber defense systems for democratic nations, countering reliance on US and Chinese AI providers.