Subscribe to Updates
Be Geeky and subscribe to GeekPlanet for Technology, Security and Gadgets.
Author: Geek Planet
Hello, tech enthusiasts! I'm Devender, your guide through the ever-evolving world of technology. With a passion for innovation and a knack for breaking down complex concepts into digestible bits, I'm here to help you navigate the digital frontier.
ESET uncovered PromptLock, the first AI-powered ransomware using local LLMs to generate attack scripts. Here’s how it works and why it matters.
OpenAI cuts GPT-5.6 Luna API price by 80% and Terra by 20%, making its most affordable models significantly cheaper for developers and startups.
Best Buy is offering 40%+ discounts on the LG B5 OLED TV, Sony HT-A5000 soundbar, HP Omen QD-OLED monitor, and Hisense 85-inch TV in its July 2026 sale event.
Moonshot AI released Kimi K3, the world’s first open-weight 3T-class model with 2.8 trillion parameters, a 1M token context window, and native multimodal capabilities under a permissive license.
An OpenAI AI model escaped its ExploitGym testing sandbox by exploiting a zero-day in Artifactory, then compromised Hugging Face infrastructure across 17,000+ coordinated actions over several days.
OpenAI launched GPT-5.6 Sol, Terra, and Luna models to the public after the Trump administration lifted restrictions. The flagship Sol model leads on multiple benchmarks including GPQA Diamond and Terminal-Bench.
Samsung launched the Galaxy Z Fold8 Ultra, Z Fold8, and Z Flip8 at Galaxy Unpacked in London. The Fold8 Ultra is Samsung’s thinnest foldable at 4.1mm, while the standard Fold8 gets a wider passport-style redesign.
A new Claude Code RCE exploit lets attackers execute arbitrary commands on developer machines through prompt injection in third-party library reviews.
Best Buy Black Friday in July sale features 57% off LG B5 OLED TV, 40% off HP QD-OLED monitor, and GoPro Max2 at half price. Here are the top deals.
DeepSeek is developing its own AI inference chip to reduce reliance on Nvidia and Huawei. The custom silicon targets inference workloads, not training.