AI Has Redefined Software Risk - Learn How Security Teams Can Update Their PlaybookRegister Now
The Week in Security: NuGet hit by typosquatting, fake ChatGPT plug-in hijacks Facebook accounts

The Week in Security: NuGet hit by typosquatting, fake ChatGPT plug-in hijacks Facebook accounts

Welcome to the latest edition of The Week in Security, which brings you the newest headlines from both the world and our team across the full stack of security: application security, cybersecurity, and beyond.

The Week in Security: NuGet hit by typosquatting, fake ChatGPT plug-in hijacks Facebook accounts
The Week in Security: YoroTrooper steals credentials in Europe, AI-created videos spread malware

The Week in Security: YoroTrooper steals credentials in Europe, AI-created videos spread malware

Welcome to the latest edition of The Week in Security, which brings you the newest headlines from both the world and our team across the full stack of security: application security, cybersecurity, and beyond.

The Week in Security: YoroTrooper steals credentials in Europe, AI-created videos spread malware
pulling back curtain on big eye looking through keyhole

Introducing New Secrets Management Capabilities For Mitigating Software Supply Chain Risk

ReversingLabs is giving dev and app sec teams something new: broader visibility into software supply chain risks — and data-driven prioritization to automatically suppress third-party secrets and other false positive results.

Introducing New Secrets Management Capabilities For Mitigating Software Supply Chain Risk
The Week in Security: Lazarus attacks same South Korean entity twice, use of hard-coded secrets is up

The Week in Security: Lazarus attacks same South Korean entity twice, use of hard-coded secrets is up

Welcome to the latest edition of The Week in Security, which brings you the newest headlines from both the world and our team across the full stack of security: application security, cybersecurity, and beyond.

The Week in Security: Lazarus attacks same South Korean entity twice, use of hard-coded secrets is up
Back to Top