Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Your Bedrock agent running on EKS receives a prompt through your RAG pipeline. CloudTrail logs it as a normal bedrock:InvokeModel event—status 200, authorized IAM role, expected endpoint. But inside the container, the agent’s response triggers a tool call that spawns curl to an external IP, exfiltrating the context window. GuardDuty doesn’t flag it because the connection routes through a permitted VPC endpoint. You open your AWS console and see a healthy API call.

Before the AI spectacle of RSA arrives, let’s talk about what actually keeps regulated organizations secure RSA is only weeks away. And if you’ve been paying any attention to the pre-conference buzz, or if you work in technology generally, you already know what it’s going to feel like walking that floor: artificial intelligence, everywhere, in everything. AI-powered detection. Autonomous response. Agentic security copilots in everything from threat monitoring to your morning coffee.

As organizations continue to scale their AWS environments, security teams face increasing challenges in detecting cloud-native threats such as compromised credentials, misused APIs, container breaches, and malicious workload behavior. Traditional perimeter-based controls and legacy endpoint tools are often insufficient in dynamic, cloud-first architectures. AWS GuardDuty provides native,intelligent threat detection for AWS environments.

Incident response in the cloud is derailed not by a lack of skill, but by a lack of visibility. Security teams frequently discover critical blind spots only after an incident is already underway, leading to delayed containment, inaccurate attribution, and incomplete forensic analysis. This report walks through six realistic, real-world inspired scenarios where missing log sources prevented effective investigations.

The modern workplace has a new “system of record,” and it isn’t email. Today, approvals, incident coordination, customer escalations, vendor conversations, quick file shares, and “can you grant access?” requests happen in Slack channels, Teams chats, and Google Chat spaces, often at a pace that makes formal controls feel optional.

A useful way to evaluate a modern identity security platform is to look at three core pillars: strong authentication and access controls, Privileged Access Management (PAM) that reduces standing privilege and secure credentials and secrets management with continuous visibility. In this blog, we’ll break down what those pillars mean, how they work together and how to evaluate and roll them out without creating gaps or friction as you scale automation and agent-driven workflows.

Every new AI feature that ships into a platform your employees already use is a security question your stack probably can't answer yet. It sounds like hyperbole, but it's the structural reality of how AI adoption works in 2026. A recent update to WhatsApp is a useful illustration of why.

At Cloudflare, our mission is to help build a better Internet. Usually, that means rolling out new services to our millions of users or defending the web against the world’s largest cyber attacks. But sometimes, building a better Internet requires us to stand up against laws or regulations that threaten its fundamental architecture.

In July 2024, DigiCert discovered they’d been issuing certificates with improper domain validation for five years. They gave customers 24 hours to replace 83,000 certificates. CISA issued an emergency alert. Critical infrastructure operators couldn’t meet the deadline. Some customers sued. That’s what mass revocation looks like in practice. The CA finds a compliance problem, the clock starts, and everyone scrambles. ACME Renewal Information (ARI) is the fix.

Credential abuse, exploitation of vulnerabilities, or phishing were the initial access vectors in 61% of breaches in 2025, according to Verizon’s 2025 Data Breach Investigation Report. While new threats present fresh challenges to security teams, reports like this highlight that cybercriminals still favor well-established attack methods and exploit familiar weaknesses.