Biztonsági szemle

A researcher-developed framework could enable attackers to conduct real-time conversations using simulated audio to compromise organizations and extract sensitive information.

The Internet of Things (IoT) has made everything more interconnected than ever, but an important US government security initiative is stuck in limbo even as threat actors step up attacks on everything from medical gear to printers.

The first known malicious MCP server is an AI integration tool that automatically sends email such as those related to password resets, account confirmations, security alerts, invoices, and receipts to threat actors.

Akira ransomware actors are currently targeting SonicWall firewall customers vulnerable to a bug discovered last year.

Attackers impersonate the National Police of Ukraine to deploy Amatera Stealer and PureMiner, using malicious Scalable Vector Graphics to trick victims.

A Nemzeti Kiberbiztonsági Intézet riasztást ad ki a Cisco ASA/FTD tűzfalakat érintő, több – közülük kettő aktívan kihasznált – sérülékenység, valamint a sebezhető eszközök elleni aktív kibertámadási hullám kapcsán. A kampány célzottan az ASA/FTD...

Three international vehicle manufacturers have fallen to supply chain cyberattacks in the past month alone.

Security researchers say multiple threat groups, including Iran's Charming Kitten APT offshoot Subtle Snail, are deploying malware with code-signing certificates from the Houston-based company.

The world's most-popular sports contest starts in June 2026 across 16 venues in three countries: Securing the event infrastructure from cyber threats will require massive collaboration.

Patch now: Cisco recently disclosed four actively exploited zero-days affecting millions of devices, including three targeted by a nation-state actor previously discovered to be behind the "ArcaneDoor" campaign.

The China-linked cyber-espionage group UNC5221 is compromising network appliances that cannot run traditional EDR agents to deploy new versions of the "Brickstorm" backdoor.

Yet again researchers have uncovered an opportunity (dubbed "ForcedLeak" for indirect prompt injection against autonomous agents lacking sufficient security controls — but this time the risk involves PII, corporate secrets, physical location data...