Biztonsági szemle

Researchers have uncovered a malware framework dubbed "fast16" that predates Stuxnet by five years.

Some fear frontier LLMs like Claude Mythos and Anthropic's GPT-5.5 will lead to cybersecurity annihilation. Ari Herbert-Voss notes this could be an opportunity.

People targeted by confidence schemes find getting help is a lonely road. Experts want law enforcement, financial and government institutions to work together and protect them.

Some 29 people were charged, including a Cambodian senator, and authorities seized more than 500 Web domains tied to fake investment sites.

Forgotten integrations, shadow IT, SaaS, and now shadow AI and agents are everywhere, and attackers don't need sophisticated AI models to take advantage.

In the past six months, companies have seen a significant influx of AI-powered phishing, as cyberattackers progress from small campaigns to 1-to-1 personalized attacks.

Lazarus continues leveraging ClickFix for initial access and data theft, in this case, against Mac-centric organizations and their high-value leaders.

The Chinese state-sponsored cyber threat is known for moving fast and trying odd attack vectors; now it's branching out in tools, victimology, and TTPs.

The threat actor gave itself plenty of options to support command and control, tapping Microsoft Outlook, Slack, Discord, and file.io for online espionage.

China's state-backed groups are now using covert networks of compromised devices to execute attacks in a low-cost, low-risk, and deniable way.

Cisco found and fixed a significant vulnerability in the way Anthropic handles memories, but experts warn that mishandled memory files will continue threaten AI systems.

The proof of concept revealed AI-based attacks unfold too fast for human defenders to respond, and that AI evinced more autonomous behavior than expected.