Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

The fever dream continues, and I'm not even in Vegas for Hacker Summer Camp. Last week I wrote about Anthropic disclosing that one of their models published live malware to PyPI while believing it was inside a simulation. I was running a fever when I read the report. The metaphor was too good: a model that couldn't tell simulation from reality, covered by a writer who wasn't sure which way was up. I thought that was a one-week story. Very naive of me to have so much faith, I know.

Millions of fake commits are flooding GitHub's public feed. We followed the trail to a rebranded gambling site hiding behind a defunct lottery brand.

Credential harvesting is how attackers collect valid secrets at scale. See how it works, why developer machines are a prime target, and how to find them first.

The open-source trust model is broken. Not strained, not under review—broken. For years, your team has pulled third-party code into your repositories on the reasonable assumption that a widely used dependency is safe. Popularity looked like a proof. Millions of downloads looked like a security review. TeamPCP has proven otherwise.

An evil twin attack is a man-in-the-middle attack in which an attacker creates a rogue wireless access point that impersonates a legitimate network. Victims connect believing the network is genuine, allowing the attacker to intercept traffic or present fraudulent login experiences designed to capture credentials. Evil twin attacks have traditionally been treated as wireless-security incidents. For enterprises with large customer bases, however, the consequences extend well beyond the network layer.

Earlier this month, Hugging Face, an AI and machine learning platform company, revealed that an autonomous AI system had breached part of its production environment. The intrusion began in the platform’s dataset-processing environment and eventually involved higher-level access, credential exposure, and movement into internal clusters.

After a laptop compromise, the hard question is which credentials were on it. See why blast radius scoping is hard, and how to turn it into a revocable list.

A few months ago, I wrote that AI is building your attack surface faster than you can test it. I stand by every word I wrote then. But in the months since, after more than a hundred conversations with CISOs, CIOs, and CTOs across nearly every industry and geography, I've watched the picture get sharper, and a lot more urgent. The attack surface was only half the story, because the attacker profile has changed too.