JADEPUFFER: The First Fully Autonomous AI Ransomware Attack — And Why Your Documents Are the Target

Key Takeaways • JADEPUFFER is the first confirmed AI agent to execute a full ransomware attack with zero human involvement. • The attack exploited CVE-2025-3248, a critical Langflow vulnerability (CVSS 9.8), to harvest cloud credentials and pivot to a production database. • Recovery was structurally impossible: the encryption key was never saved, even after ransom was paid. • AI-driven ransomware collapses attack timelines from weeks to hours, with no human operator required. • FED...

Key Takeaways • JADEPUFFER is the first confirmed AI agent to execute a full ransomware attack with zero human involvement. • The attack exploited...

Key Takeaways • On September 11, 2026, EU CRA vulnerability reporting obligations go live — covering both new and legacy products. • Manufacturers must submit a 24-hour early warning and a 72-hour full notification when an actively exploited vulnerabilities is identified. • Non-compliance carries fines of up to €15 million or 2.5% of global annual turnover, whichever is higher. • The obligations apply to all products with digital elements on the EU market —...

Key Takeaways • On September 11, 2026, EU CRA vulnerability reporting obligations go live — covering both new and legacy products. • Manufacturers must...

Screen capture is the overlooked blind spot in financial services, the one exit a bank’s perimeter was never built to watch. Every financial institution’s worst insider scenario recently played out somewhere else first. In late 2025, a major technology company discovered that one of its own employees, someone with legitimate access to internal systems, had been capturing screenshots of sensitive internal consoles and passing them to a criminal hacking group, which posted them publicly....

Screen capture is the overlooked blind spot in financial services, the one exit a bank’s perimeter was never built to watch. Every financial institution’s...

Key Takeaways • ShinyHunters breached three major healthcare organizations in a three-month campaign — exposing over 9 million records and 8.8 TB of patient data. • Healthcare breaches cost an average of $7.42 million per incident — the highest of any industry for 14 consecutive years. • 67% of healthcare organizations face ransomware attacks annually, with 20%+ reporting increased patient mortality following an attack. • EDRM renders exfiltrated files unreadable — eliminating the attacker’s...

Key Takeaways • ShinyHunters breached three major healthcare organizations in a three-month campaign — exposing over 9 million records and 8.8 TB of patient...

Data security has a sequencing problem. Most organizations invest heavily in encryption, access controls, and data loss prevention—before they know where their sensitive data actually lives. Those controls are not wrong. However, they are blind. They protect what is already mapped, not what is hidden. The result is a data visibility gap that grows with every cloud migration, SaaS adoption, and employee AI prompt. Data you cannot see cannot be protected. In 2026, that...

Data security has a sequencing problem. Most organizations invest heavily in encryption, access controls, and data loss prevention—before they know where their sensitive data...

Key Takeaways • Data exfiltration spikes 720% in the 24 hours before a layoff — but suspicious activity often starts up to six months earlier. • 59% of departing employees take confidential data with them, with personal cloud storage now the #1 exfiltration channel. • Traditional DLP tools cannot protect files after they leave the corporate network — once a document reaches a personal device, DLP has no visibility. • Persistent document-level encryption renders...

Key Takeaways • Data exfiltration spikes 720% in the 24 hours before a layoff — but suspicious activity often starts up to six months...

Key Takeaways • In May 2026, attackers stole 275 million records from an education platform — the largest education data breach on record. • Platform-level encryption did not protect the content. Once attackers reach the application layer, data is fully readable. • Paying the ransom for “data destruction” is unverifiable — and not a security strategy. • The fix is data-centric security: encrypt the content itself, not just the container. A criminal group...

Key Takeaways • In May 2026, attackers stole 275 million records from an education platform — the largest education data breach on record. • ...

May 14, 2026
Blog, PQC

The Mythos Shock: When AI Becomes the Threat In April 2026, Anthropic revealed that its next-generation model, Claude Mythos Preview, had autonomously discovered thousands of high-risk vulnerabilities across major operating systems and web browsers. Due to its potential risks, access to the model was not made public but instead restricted to select partners under Project Glasswing. Somewhere in a government data center, a server rack is quietly archiving petabytes of encrypted network traffic—financial records,...

The Mythos Shock: When AI Becomes the Threat In April 2026, Anthropic revealed that its next-generation model, Claude Mythos Preview, had autonomously discovered thousands...

The moment a company enters a merger or acquisition process, it opens a document repository containing almost everything an adversary could want: unreported financials, IP portfolios, regulatory filings, strategic plans, and personnel records. For weeks or months, those documents are deliberately shared with counterparties and advisors outside the organization’s normal security perimeter. That window of deliberate openness is one of the most dangerous periods in any organization’s security calendar. And attackers know exactly when...

The moment a company enters a merger or acquisition process, it opens a document repository containing almost everything an adversary could want: unreported financials,...

The Stakes Have Changed for Defense Contractors For years, cybersecurity compliance in the defense industrial base (DIB) was largely a matter of self-attestation. Companies certified their own security posture, and enforcement was inconsistent. That era is ending. The Cybersecurity Maturity Model Certification (CMMC) 2.0 framework, now entering full enforcement as part of the U.S. Department of Defense (DOD)’s acquisition process, fundamentally changes the equation. Third-party assessments, stricter requirements for handling Controlled Unclassified Information (CUI),...

The Stakes Have Changed for Defense Contractors For years, cybersecurity compliance in the defense industrial base (DIB) was largely a matter of self-attestation. Companies...

AI Data Platform

Solution

Product

Solution

Product

Enterprise LLM

AI Consulting

Unstructured Data Management

Solution

Product

Data-Centric Collaboration

Data-Centric Messaging

AI

Solution

Product

Discovery & Classification

Data Security Posture Management (DSPM)

Enterprise DRM (EDRM)

Solution

Product

Endpoint Protection

Secure Collaboration

Enterprise Backup and Recovery (EBR)

Data Detection and Response (DDR)

Blog