darkreading

Microsoft previewed the Sentinel security graph and MCP server at its annual Microsoft Secure virtual event earlier this month.

The vast majority of organizations are encountering AI-augmented threats, but remain confident in their defenses, despite inadequate detection investment and more than half falling to successful attacks.

Dark Reading is looking for leading industry experts with a point of view they want to share with the rest of the cybersecurity community for our new Commentary section.

While GitHub has advanced protections for its built-in AI agent, a researcher came up with a creative proof-of-concept (PoC) attack for exfiltrating code and secrets via Copilot.

SonicWall said a breach it disclosed last month affected firewall configuration files for all customers who have used SonicWall’s cloud backup service — up from its previous 5% estimate.

Marshall Erwin shares how crisis leadership shaped his path from CIA analyst to the US Congress to protecting global Web traffic at Fastly.

Transcription applications are joining your online meetings. Here's how to create policies for ensuring compliance and security of your information.

New encryption, wiper, and cryptocurrency-stealing capabilities make the evolving ransomware-as-a-service operation more dangerous than ever.

The campaign is the latest by BatShadow, one of a growing number of cybercrime groups operating out of Vietnam.

Crimson Collective, which recently breached the GitLab instance of Red Hat Consulting, has teamed up with the notorious cybercriminal collective.

The three extortion gangs also invited other e-crime attackers to join their collaboration to share attack information and resources, in the wake of LockBit 5.0 being released.

Patch now: A bug (CVE-2025-53967) in a third-party option for connecting Figma to agentic AI can lead to remote code execution (RCE).

A threat actor is putting a spin on classic remote monitoring and management (RMM) attacks, using a Chinese open source tool instead.

Wanna work for a hot brand? Cyberattackers continue to evolve lures for job seekers in an impersonation campaign aimed at stealing résumés from social media pros.

A ransomware last week left the Asahi brewery in Japan struggling to take orders and deliver its products domestically, as manufacturers become a favored target.

Researchers report an increase in the use of hidden content in spam and malicious email to confuse filters and other security mechanisms.

In a recent poll, readers shared how they're using vibe coding in AppDev (if they are at all). While some found success, others found the risks too great.

Researchers say exploitation of CVE-2025-10035 requires a private key, and it's unclear how Storm-1175 threat actors pulled this off.

A 13-year-old flaw with a CVSS score of 10 in the popular data storage service allows for full host takeover, and more than 300k instances are currently exposed.

A threat actor purporting to be from the Libyan Navy's Office of Protocol targeted Brazil's military earlier this year using the rare tactic.