Biztonsági szemle

Teams are putting Claude Enterprise to work fast, and not just for chat. Claude, Claude Code, and Claude Cowork act on behalf of users. They read documents, call tools, and run tasks. That makes the prompt an attack surface. A jailbreak can try to..

New research shows how attackers can use security alerts and blocked events to manipulate and hijack AI agents.

Attacks targeting water systems just keep flowing across a dozen states, against ill-secured, Internet-exposed PLCs.

The maximum-severity vulnerability, which still has no CVE, allows malicious, remote administrator access to the business-analytics platform and its downstream users.

It's time to turn from CVSS-backed patching to choke-point patching focused on breaking chains to critical assets.

Sophisticated iPhone exploit chains previously limited to nation-states are spreading far and wide to organized cybercrime groups.

A public policy expert mapped global cybercrime laws to develop a five-point framework for protecting ethical hackers and good-faith security research.

Use Cisco's beta Meraki and Catalyst Center MCP servers to bring supported, read-only network data into the AI agents and applications you build.

The crime solver wore disguises, spied on targets, and built intelligence networks long before modern-day tactics emerged. He has lessons for today’s ethical- and nonethical-hat hackers.

Cisco is proud to announce a landmark achievement: Secure Email Threat Defense has officially achieved FedRAMP Class D (High) certification!

Auth is the other pillar, with a sneakier quantum deadline: a live signature need only resist forgery until it's verified, but long-lived trust anchors and slow PKI migration mean roots must go quantum-safe early. We meet ML-DSA and SLH-DSA

A study of more than 6,000 patches found that even working patches can introduce new bugs, break something else, or are open to bypass.