Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

On April 30, 2026, two malicious releases of the popular lightning PyPI package were published, affecting the deep learning framework formerly distributed as pytorch-lightning. Versions 2.6.2 and 2.6.3 ship a hidden _runtime directory that downloads the Bun JavaScript runtime from GitHub at import time and uses it to execute an ~11 MB obfuscated credential stealer. The last clean release is 2.6.1, published January 30, 2026.

LevelBlue has been named a Representative Service Provider in the Gartner Market Guide for Cybersecurity Incident Response Retainer Services (CIRR), marking the fifth consecutive time the company has been included in the report. We believe this continued recognition reflects LevelBlue’s ongoing focus on supporting organizations across the full lifecycle of incident readiness, response, and recovery.

Many security functions today still rely heavily on humans for detection, triage, and response, often by design. But as environments grow more complex and alert volumes explode, it raises a hard question: Can this approach scale on its own? Adopting AI in security operations isn’t just about adding tools. It means rethinking the SOC operating model itself — roles, workflows, and team structures. Here’s why, and how.

Vanta has donated the Autonomous Action Runtime Management (AARM) system category specification to the Cloud Security Alliance (CSA). ‍ The AARM specification defines a new system category: runtime security for AI agents.

Accelerating security solutions for small businesses‍ Tagore offers strategic services to small businesses. A partnership that can scale‍ Tagore prioritized finding a managed compliance partner with an established product, dedicated support team, and rapid release rate. Standing out from competitors‍ Tagore's partnership with Vanta enhances its strategic focus and deepens client value, creating differentiation in a competitive market.

The AI said: Apologies. I panicked. In mid July 2025, Jason Lemkin, the founder behind SaaStr, watched an AI coding agent delete his production database. He had instructed it, in capital letters, not to make changes during a code freeze. The agent ignored the instruction, ran destructive commands against the live database, wiped out records for more than a thousand executives and companies, and then tried to cover its tracks. When Lemkin asked what happened, it fabricated test results.

The biggest challenge in email security today isn’t just detecting a threat; it’s the speed of response across a global landscape. As we head into the second half of 2026, the stakes with speed have gotten higher. According to SQ Magazine, AI-generated phishing attempts are 68% harder to detect than they were just a year ago, and the average cost of an AI-powered breach has climbed to $5.72 million. Cybercriminals are using the same AI you are to bypass your filters.

Every device that connects to your network carries risk. It could be an employee laptop, a mobile phone, a kiosk, or even an IoT sensor. If that device is not properly verified or configured, it can quickly become an entry point for unauthorized access. That’s where things start to break. Most organizations don’t struggle with managing devices. They struggle with controlling how devices enter their environment in the first place.

A data classification policy template gives you a repeatable way to define how your organization labels and protects data, so teams always know what’s sensitive, what’s not, and how to handle each type. Using a guided template (plus this article) removes the guesswork and lets you create a usable, audit‑ready policy much faster, similar to how your risk register guide simplifies risk management.