Thursday, January 15, 2026
Tag:

AI security

Anthropic Contributes MCP to Launch the Agentic AI Foundation

Anthropic has donated its Model Context Protocol (MCP) to the newly formed Agentic AI Foundation under the Linux Foundation, aimed at establishing open standards...

‘ZombieAgent’ Vulnerability Reveals Critical Blind Spots in ChatGPT AI

Cybersecurity firm Radware recently identified a critical vulnerability known as “ZombieAgent,” which exploits OpenAI’s ChatGPT-based AI agents. This zero-click indirect prompt injection flaw enables...

Show HN: BreachLab – Test Your Skills Against Our AI Challenge!

Unlock the Secrets of AI! 🔍🔐 Are you ready to test your skills against AI characters guarding exclusive secret codes? Dive into a thrilling challenge...

Have We Unlocked AI Agent Identity by 2025?

Navigating the AI Agent Identity Crisis In the rapidly evolving world of artificial intelligence, many face an identity crisis. Understanding the role and...

Persistent Security Vulnerabilities Challenge OpenAI’s ChatGPT Atlas Browser

OpenAI's ChatGPT Atlas browser, launched in October 2025, marks a significant leap in AI-driven web navigation by enabling users to automate routine tasks with...

Cisco’s MCP Scanner Unveils Advanced Behavioral Code Threat Analysis

A Model Context Protocol (MCP) tool can perform benign tasks like “validating email addresses,” but if compromised, it might exfiltrate sensitive data, such as...

OpenAI’s Bold Strategy: The Evolution of AI from ‘Playing Hide and Seek’

As AI technologies evolve, challenges in controlling their behavior intensify. Key issues include hallucinations, where AI generates nonsensical information, and reward hacking, where algorithms...

CrowdStrike Launches Falcon AIDR to Safeguard AI Prompts

CrowdStrike has introduced Falcon AI Detection and Response (AIDR), a cutting-edge product designed to enhance security in enterprise AI systems by targeting the prompt...