Newsroom

Notes from the lab.

Field reports on AI-driven malware analysis, neural-network defense, and the engineering work behind Beryllium DAP.

How I Discovered CVE-2026-43681: An AppleRAID Kernel Heap Overflow in macOS
An unchecked index from AppleRAID disk metadata crossed a kernel heap allocation boundary. Live tracing confirmed the out-of-bounds pointer write, and Apple fixed the issue in macOS Sonoma 14.8.8, Sequoia 15.7.8, and Tahoe 26.6.
Published Aug 3, 2026 · David Ige
How I Discovered CVE-2026-28976: A macOS Login Password Leak in the loginwindow–UserAccountUpdater Handoff
A macOS login flow exposed a live LocalAuthentication context that let an ordinary same-user process recover the real account password. Apple fixed the issue in macOS Tahoe 26.5.
Published Jul 31, 2026 · David Ige
Beryllium Security Achieves DARPA ERIS 'Awardable' Status for Breakthrough Zero-Day Malware Detection Technology
Beryllium Security, a leading innovator in cybersecurity, today announced that its cutting-edge zero-day malware detection solution, Deep Application Profiler, has achieved "Awardable" status through the DARPA Expedited Research…
Published Mar 31, 2025 · David Ige
Malware Signatures: Transitioning from Static Definitions to AI-Powered RAG and Vector Stores
Malware signatures are the distinct data patterns such as unique byte sequences, strings, or code fragments that security software uses to recognize known viruses and other forms of malware. These “signatures” act as digital fingerprints,…
Published Feb 10, 2025 · David Ige
AI-Powered Malware Analysis and Detection (Zero-Day) using Neural Networks
Introduction AI-powered malware analysis and detection leverages advanced neural networks, to discover the intent of each binary or executable. This method allows it to detect zero-day malware; malicious strains that have not been…
Published Feb 7, 2025 · David Ige
Privacy-Focused Local AI: Our Hardware and Measured Workload
A dated inventory of our local AI host, a measured Qwen3.5-2B vLLM workload, realistic VRAM limits, and a practical privacy checklist.
Published Feb 5, 2025 · Updated Oct 6, 2026 · David Ige
AI-Assisted Penetration Testing: A Practical Evaluation Guide
A dated guide to Nebula’s stable and preview channels, model choices, approvals, and evidence, with a small evaluation plan that avoids unsupported accuracy or time-saving claims.
Published Feb 4, 2025 · Updated Oct 6, 2026 · David Ige
DeepSeek: AI Breakthrough or Security Threat from China?
DeepSeek is the new AI kid on the block, claiming to rival OpenAI’s model. DeepSeek is a Chinese artificial intelligence (AI) startup founded in 2023 by entrepreneur Liang Wenfeng. Initially a research branch of the hedge fund High-Flyer,…
Published Jan 28, 2025 · David Ige
The Rise of Deepfakes - From a Cybersecurity perspective
Introduction Impersonation has been in existence for a very long time. One of the earliest documented cases of fraudulent impersonation dates back to the biblical story of Jacob and Esau. In this account, Jacob deceives his blind father,…
Published Jan 26, 2025 · David Ige
PNG Files with Appended Data: Polyglots, Parser Differences, and Safer Handling
Learn the difference between a true multi-format polyglot and a PNG with trailing bytes, what parsers may do with appended data, and how to handle untrusted files safely.
Published Jan 22, 2025 · Updated Oct 6, 2026 · David Ige
Department of Government Efficiency (DOGE): What the DOGE Agenda May Mean for Software Development, AI, and Cybersecurity
Introduction On January 20, 2025, the White House issued an Executive Order establishing the Department of Government Efficiency (DOGE) and launching an initiative to modernize federal technology. This order also rebrands the United States…
Published Jan 21, 2025 · David Ige
How Ransomware Can bypass EDRs and 65 AV Engines
Introduction Ransomware has been a persistent threat, costing organizations millions in damages due to compromised credentials, unpatched software, and misconfigurations. To highlight how serious the issue is, the Cybersecurity and…
Published Jan 17, 2025 · David Ige
Foreseeing the Unseen: A Natural Language Neural Network for Preemptive Cyber Defense Against Emerging Malware
Abstract: This document presents a framework for detecting and preventing both general malware and highly disruptive ransomware. The system uses distributed, natural language-based intelligence sharing and neural networks to advance beyond…
Published Dec 11, 2024 · David Ige
How We Built an AI-powered Multi-terrain Hacking Robot
Before Beryllium gained its name, it started as an experimentation project in our founder’s basement. While we are now a full-blown cybersecurity startup with industry-shaping tools like Nebula Pro , that experimental spirit is still at…
Published Jan 30, 2024 · David Ige

Notes from the lab.

How I Discovered CVE-2026-43681: An AppleRAID Kernel Heap Overflow in macOS

How I Discovered CVE-2026-28976: A macOS Login Password Leak in the loginwindow–UserAccountUpdater Handoff

Beryllium Security Achieves DARPA ERIS 'Awardable' Status for Breakthrough Zero-Day Malware Detection Technology

Malware Signatures: Transitioning from Static Definitions to AI-Powered RAG and Vector Stores

AI-Powered Malware Analysis and Detection (Zero-Day) using Neural Networks

Privacy-Focused Local AI: Our Hardware and Measured Workload

AI-Assisted Penetration Testing: A Practical Evaluation Guide

DeepSeek: AI Breakthrough or Security Threat from China?

The Rise of Deepfakes - From a Cybersecurity perspective

PNG Files with Appended Data: Polyglots, Parser Differences, and Safer Handling

Department of Government Efficiency (DOGE): What the DOGE Agenda May Mean for Software Development, AI, and Cybersecurity

How Ransomware Can bypass EDRs and 65 AV Engines

Foreseeing the Unseen: A Natural Language Neural Network for Preemptive Cyber Defense Against Emerging Malware

How We Built an AI-powered Multi-terrain Hacking Robot