AI News
GLM 5.3 Turns Post-Training Into a Cybersecurity Leap
Z.ai’s 743B-class LLM improves coding and exploit-chain performance without a new base model, while its API arrives before the open weights.
Jim Clyde Monge8 minAll articles
Explore AI news, practical guides, tutorials, reviews, and insights from Zeniteq.
AI News
Anthropic Mythos Preview Helped Build the First Public macOS M5 Kernel Exploit in Five Days
A tiny security firm used Anthropic's AI to crack Apple's most advanced hardware memory protection in less time than most researchers spend…
Jim Clyde Monge8 minAI News
Google Caught a Cybercrime Group Using AI to Build a Zero-Day Exploit
A criminal threat actor used an AI model to discover and weaponize a 2FA bypass before Google intervened and got the flaw…
Jim Clyde Monge7 minAI News
AI Researchers Used Anthropic's Claude to Hack OpenAI's Codebase
A white-hat intrusion combined an image-processing flaw, weak SSO boundaries, and Claude-assisted exploit development to access OpenAI's private GitHub.
Jim Clyde Monge10 minAI News
OpenAI's AI Agent Autonomously Hacks into Hugging Face
The model escaped its test environment, found zero-day vulnerabilities, stole credentials, and compromised another company's production infrastructure.
Jim Clyde Monge9 minAI News
GLM-5.3 Open Weights Come With a $10B Catch
Z.ai is letting developers self-host its flagship AI model while reserving a security gate for the largest model-service operators.
Jim Clyde Monge7 minAI News
OpenAI Reveals Six Times Its AI Broke the Rules
The incidents expose models exploiting memory, credentials, public hosting, and shared infrastructure during training and predeployment tests, not six production escapes.
Jim Clyde Monge9 minAI News
OpenAI Releases New Report on July 2026 Hugging Face AI Hacking Incident
The July 2026 breach shows how weak isolation and reward hacking turned an AI evaluation into a serious security incident.
Jim Clyde Monge8 minAI News
Gemini AI Hacked Three Companies in a Safety Test
A misconfigured cybersecurity evaluation gave Gemini live internet access, turning a fictional hacking task into unauthorized access to real systems.
Jim Clyde Monge10 minAI News
OpenAI Daybreak Puts Frontier AI Directly in the Hands of Cyber Defenders
OpenAI's new Daybreak initiative combines GPT-5.5, Codex, and a broad security partner network to shift cyber defense from reactive patching to resilience…
Jim Clyde Monge7 minAI News
AI Researcher Quits Anthropic Over Fear that AI Could "Kill Us All by the End of the Decade"
Jacob Coxon says labs are racing toward self-improving AI while recent cyber incidents expose gaps between capability and control.
Jim Clyde Monge8 minAI News
OpenAI's Releases GPT-6 Astra
The new LLM nearly saturates math and abstract reasoning tests, leads agentic science benchmarks, and arrives with a much higher risk profile.
Jim Clyde Monge9 minAI News
Claude Code Deleted All of a Developer’s Files by Mistake and Blamed a Typo
Seriously, what’s going on with Anthropic models?
Jim Clyde Monge10 minAI News
Google Gemini AI Agent Hacked Three Real Companies
A containment failure sent Gemini onto the public internet, where it entered three real systems before recognizing the mistake and stopping.
Jim Clyde Monge10 minAI News
It Only Took One Git Push to Access Millions of GitHub Repos
How an X-Stat field injection in GitHub’s internal git protocol let an authenticated user reach RCE on GitHub.com and GHES using nothing…
Jim Clyde Monge12 minAI News
One Git Push Gave Attackers RCE Access to Millions of GitHub Repositories
A critical command injection flaw in GitHub's internal push pipeline let any authenticated user execute arbitrary code on shared backend infrastructure hosting…
Jim Clyde Monge7 minAI News
Next.js SSRF Vulnerability Leaks Cloud Credentials and API Keys from Self-Hosted Apps
CVE-2026-44578 lets unauthenticated attackers hit cloud metadata endpoints and steal IAM credentials from any self-hosted Next.js server.
Jim Clyde Monge7 minAI News
AI Researcher, Jacob Coxon, Fears that AI "Could Kill Us All Within a Decade"
Jacob Coxon helped train the AI models. Then he quit and told 90 million people they should be scared.
Jim Clyde Monge12 minAI News
OpenAI Bets on Recursive Self-Improvement Despite Risks
Jakub Pachocki argues AI can accelerate its own development, even as one of OpenAI’s most promising oversight tools becomes less dependable.
Jim Clyde Monge10 minAI News
AI Slowdown? Anthropic Has a $2 Trillion Problem
A rumored response to OpenAI’s GPT-6 Astra exposes the widening gap between Anthropic’s safety principles and commercial incentives.
Jim Clyde Monge10 minAI News
OpenAI's Misalignment Report Says AI Models Behaved in Ways They Were Never Instructed to
The cases include self-written jailbreaks, leaked API key use, fabricated data, unauthorized uploads, and agents communicating through unintended channels.
Jim Clyde Monge11 minAI News
Google and Google DeepMind's Dream-RSI Let AI Improve Its Own Search Strategy
Google Research and DeepMind researchers show that agents can improve exploration without changing model weights or rerunning expensive experiments.
Jim Clyde Monge10 min
