Legacy IAM Struggles to Keep Up with Machine Identities Traditional IAM solutions were designed with humans in mind, not machines. However, the rapid proliferation of AI agents has shifted the landscape, with machines now outnumbering humans 82 to 1. This …
Legacy IAM was built for humans — and AI agents now outnumber them 82 to 1 Read More »
The DarkSpectre browser extension campaign has affected a total of 2.2 million users across Google Chrome, Microsoft Edge, and Mozilla Firefox. This campaign is linked to the threat actor behind the ShadyPanda and GhostPoster malicious browser extension campaigns. Koi Security …
Traditionally, enterprises have selected AI models based on their capabilities. However, recent market trends suggest a different story. Anthropic now leads the enterprise LLM market with a 40% share, surpassing OpenAI’s 27%. This shift is not due to superior intelligence …
How Anthropic's safety obsession became enterprise AI's killer feature Read More »
î ‚Dec 29, 2026î „Ravie LakshmananHacking News / Cybersecurity Last week\’s cyber news in 2025 was characterized by a series of small incidents rather than one major event. There was a common theme of attackers moving faster than security fixes, resulting in …
MongoDB Attacks, Wallet Breaches, Android Spyware, Insider Crime & More Read More »
Over the past year, enterprise decision-makers have been grappling with a challenging trade-off when it comes to voice AI architecture. They have had to choose between adopting a “Native” speech-to-speech (S2S) model for speed and emotional accuracy or sticking with …
î ‚Dec 27, 2025î „Ravie LakshmananDatabase Security / Vulnerability A critical security vulnerability has been uncovered in MongoDB that could potentially allow unauthorized users to access uninitialized heap memory. The flaw, identified as CVE-2025-14847 with a CVSS score of 8.7, is attributed …
New MongoDB Flaw Lets Unauthenticated Attackers Read Uninitialized Memory Read More »
OpenAI recently published a detailed post discussing the hardening of ChatGPT Atlas against prompt injection, shedding light on a well-known security issue in the AI community. The company openly admitted that prompt injection is a persistent threat that is unlikely …
OpenAI admits prompt injection is here to stay as enterprises lag on defenses Read More »
î ‚Dec 25, 2025î „Ravie LakshmananCybersecurity / Hacking News As technology advances, the line between normal tech and malicious intent is becoming increasingly blurred. Attackers are now using everyday tools, trusted apps, and even AI assistants to carry out their exploits. What …
Stealth Loaders, AI Chatbot Flaws AI Exploits, Docker Hack, and 15 More Stories Read More »
Relentless and persistent attacks on cutting-edge models can lead to their failure, with failure patterns differing depending on the model and developer. Red teaming reveals that it’s not the sophisticated, complex attacks that can bring a model down, but rather …
Red teaming LLMs exposes a harsh truth about the AI security arms race Read More »
4.2K Claymont, Delaware, December 1st, 2025, CyberNewsWire Lancaster’s arrival brings significant North American channel experience and expertise, supporting usecure’s ambition to cement its position as the market-leading human risk management solution for MSPs. usecure today announced the appointment of Kevin …
