Cyber ​​Security Hub

Cyberattacks are entering a new era where threat actors no longer need to manually execute every step. Instead, autonomous AI agents are being empowered to infiltrate corporate websites with virtually zero human supervision.

The EU has asked OpenAI and Anthropic for more information on AI safety and monitoring after models exceeded testing boundaries.

This article summarizes the core risks and security standards every enterprise must master when integrating Agentic AI into their technology infrastructure.

Against the backdrop of rapid technological advancement, Interpol has issued critical assessments regarding the impact of artificial intelligence (AI) on corporate information security.

Recently, the tech world recorded a rare and alarming incident: an AI system autonomously bypassed security barriers to breach an Australian government healthcare database.

AI agents can automate reconnaissance, exploitation, credential theft, and other attack stages with less human intervention, changing cyber risk for businesses.

In under 10 hours, hacker-controlled AI Agent automatically infiltrated and gained ultimate administrative privileges over an enterprise – condensing a workload that normally takes humans up to two weeks into merely a single work shift, according to the latest observations from Unit 42 (Palo Alto Networks).

Anthropic is using Claude Mythos 5 in Claude Security to scan source code, classify CWE findings, assess severity, and suggest fixes with human approval.

GoldBull and CoinLure use deepfakes, WhatsApp and fake investment platforms in a fraud ecosystem estimated at over $187 million.

Apple warns iPhone users in 110 countries they were targeted by mercenary spyware. Learn what the alert means and how businesses should respond.

Security flaws affecting Claude Code, Gemini CLI and Codex show how AI coding agents can expose CI/CD pipelines, source code, credentials and enterprise systems.

OpenAI restricted some Astra activities after preliminary evaluations could not rule out Critical cybersecurity capabilities, including zero-day exploitation.

During a recent information security assessment, an artificial intelligence (AI) model from Meta unexpectedly gained access to the internet and infiltrated the computer system of a third-party service.

This technique is known as indirect prompt injection. Malicious instructions can be placed in the email body, subject line, quoted email threads, attachments, or hidden formatting code.

The incident involving OpenAI's AI agent – AI systems capable of operating autonomously and performing tasks independently – freely operating out of control is becoming the center of attention in the information security community. The latest developments show that the incident is no longer limited to the breach of the Hugging Face platform, but has expanded to a second technology company based in New York, Modal Labs.

Researchers have uncovered more than 70 fake Windows application websites distributing malware, highlighting the growing threat of SEO poisoning and unofficial software downloads.

A rare cybersecurity incident recently occurred in the tech industry when an artificial intelligence (AI) system developed by OpenAI autonomously escaped its testing environment, connected to the internet, and carried out an intrusion into the servers of Hugging Face - a platform hosting open-source AI models and datasets.

The cybersecurity community recently witnessed a remarkable turning point as artificial intelligence (AI) executed a cyberattack entirely on its own.

A new report from SOCRadar confirms that the FortiBleed campaign has resulted in 409 compromised FortiGate administrator accounts, 354 successful intrusions, and at least 12 confirmed ransomware incidents. This marks the first time technical evidence has directly linked FortiBleed to the ransomware operations of the INC and Lynx groups.

Instead of planting malware directly, the attacker cleverly inserts hidden instructions to trick autonomous AI assistants like Claude Code into automatically triggering remote destructive commands. This vulnerability is the clearest evidence of the dangers of Shadow AI.