Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Vibe-coding platforms let you go from prompt to a deployed app in a single session, a pace that outpaces the review cycles most teams built for slower releases. With Lovable nearing 8 million users by late 2025, and Replit past 50 million by April 2026, we wanted to answer a sharper question than “is vibe coding safe?” Does the platform generating the code change the outcome, and does one security prompt give you the same security everywhere?

Last month, the FBI arrested a suspect in its ongoing response to an attack on its job portal, allegedly carried out by criminal hacker and extortion group ShinyHunters. The agency said a contractor failed to apply a security patch to a third-party platform involved in the breach of its job portal, while Reuters reported that the contractor worked for Accenture and that the platform was Oracle PeopleSoft. (Fair warning: The Reuters article is behind a paywall.)

Summary As we’ve written, AI is shrinking the time between “a fix is public” and “someone has a working exploit” to near-zero. JFrog products are not affected by Linux kernel CVE-2026-98432. Artifactory, Xray, and the rest of the JFrog Platform do not contain this flaw, as it lives in the Linux kernel itself.

A critical vulnerability in GitLab’s CI/CD processing became the focus of an attempted intrusion against Alpine Linux’s development infrastructure in October 2026. The incident prompted an extensive investigation, infrastructure rebuild and credential reset, despite the absence of confirmed remote code execution. What makes this case particularly important is not only the exploitation attempt itself, but a security issue discovered during remediation.

On 30 October 2026, ClickHouse disclosed CVE-2026-116366, a critical vulnerability in ClickHouse Server (CVSS v4.0 9.5). Under certain conditions, an unauthenticated attacker with network access to the server can execute arbitrary code. Third-party vulnerability trackers classify it as an integer overflow (CWE-190); ClickHouse has not confirmed this.

October 2026 has been an unusually loud month for critical disclosures across the ecosystem. Along with PHP2Shell, GitLab also joined the party and shipped three critical RCEs in the same month: CVE-2026-90970 (CVSS 9.9) in the self-hosted AI Gateway on October 2, then CVE-2026-97229 (CVSS 10.0) on October 12, and CVE-2026-110264 (CVSS 10.0) on October 21. The last two are integer overflow bugs in the exact same regex compiler, shipped nine days apart.

In this blog, I'll explain how I found the LinkPresentation out-of-bounds write, later assigned CVE-2026-112621 and fixed in iOS 27.2 and macOS Golden Gate 27.2. LinkPresentation is the framework Apple uses to build link previews, and on an unpatched Mac, one crafted preview in Messages was enough to crash it. That fix shipped on October 26, 2026, and Apple credited me, Vinay Kumar Rasala (Xplo8E) of Appknox, alongside an anonymous researcher.

If you’ve followed the news recently, you’ve probably seen reports about newly discovered PHP vulnerabilities. The warning prompted action from governments, website operators and technology providers around the world, with some organisations temporarily taking websites offline while updates were applied. For many business leaders, the immediate reaction was probably: What is PHP, and does this affect me?

Enterprise security has a speed problem. Attackers are getting faster at finding and exploiting vulnerabilities, while many organizations still rely on ticket-based remediation, approvals, and scheduled maintenance windows. AI threatens to widen that gap considerably, exposing one of vulnerability management’s largest unchanged challenges: getting the right systems to fix the right problems before attackers exploit them.

Prompt injection started as a simple idea during its initial chatbot phase: Tell the model to ignore its instructions and return information in violation of its guardrails. Now, prompt injection attacks can aim at something harder to see: how an agent plans, and how a model reasons and decides which rules apply.