Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Supply chain attacks are having a moment. Open-source malware detections jumped 73% in 2025. In the past year, the debug and chalk packages were backdoored, the tj-actions GitHub Action was compromised and pulled malicious code into thousands of pipelines, and the axios maintainer account was hijacked and used to distribute a RAT. Malicious releases also hit Zapier, ENS Domains, PostHog, and Bitwarden CLI. Every one of these attacks was preventable with controls that were available at the time.

It’s amazing how non-developers have recently been empowered to create their own apps that can even generate revenue. We’ve recently seen progress across the AI development field, from AI being successful in “greenfield code” (apps built from scratch) towards “brownfield code” (larger scale existing applications).

Pentesting made sense when releases happened every few months. A point-in-time assessment could provide an accurate picture of risk for weeks, sometimes months. Today, engineering teams ship continuously. Our State of AI in Pentesting survey of 200 CISOs and 200 engineering leaders, found that 76% deploy significant changes at least weekly, while nearly 40% deploy daily. Yet only 21% validate security on every release. That gap has consequences.

A malicious release of @injectivelabs/sdk-ts, an npm package that pulls around 50,000 weekly downloads, shipped code that records wallet mnemonics and private keys as they are derived and ships them to an attacker-controlled endpoint. The bad version, 1.20.21, was live on npm for under an hour on June 8, 2026 before the maintainer noticed and published a clean fix.

Applications using MongoDB have a common pitfall of treating the ObjectId() function as cryptographically secure. Recently, we found Rocket.Chat, an open source Slack-like application, to be a victim of this. At Aikido, we run AI Pentests on various open source applications to test our agents and identify their strengths and improvement points. During the pentest, one of the agents reported that an unauthenticated Rocket.Chat user can access any uploaded file if they know its ID.

June 10th, we announced a critical vulnerability in phpBB that lets attackers bypass authentication, now known as CVE-2026-48611. This post is a follow-up, containing technical details that explain exploit scenarios and detection methods. To get you up to speed, phpBB is an old forum software that's still being used today by various technical communities. phpBB's Site Showcase alone has over 6 million members.

For two decades, “penetration testing” has meant the same thing: once a year, you hire a firm, a human tester spends a week or two on your systems, and you get a PDF. Most compliance frameworks were written around exactly that ritual, a slow, manual, point-in-time engagement. Software doesn’t ship once a year anymore. It ships many times a day.

Last week, GitHub released self-service credential revocation for Enterprise. The feature lets organization owners cut off compromised credentials across the entire organization in one action instead of trying to track down individual tokens during an active incident. This fix was a long time coming, as the past few months have shown what happens when revocation is slow or incomplete.

npm shipped a new protection this week for its most depended-on accounts. When npm detects a sensitive action on a high-impact account, like an email swap or the use of a 2FA recovery code, it puts that account into a 72-hour read-only state and sends an alert to the previous email address. The package installs and downloads keep working as normal during this time, and the freeze lifts automatically at the end of the waiting period.

On Jun 24, 2026, the codfish/semantic-release-action GitHub Action was compromised through an imposter commit attack. An attacker force-pushed two malicious commits into the repository and repointed sixteen tags to them, including the floating major version tags v2, v3, v4, and v5. Any workflow referencing the action by one of those tags will pull and run the attacker's code on its next CI run.