Chinese Z.ai releases open source GLM-5.2, experts warn of cyberattacks
A Chinese company, Z.ai, just dropped a new open-source AI model called GLM-5.2, and it's making waves.
The model is as good as Anthropic's top tools at spotting software vulnerabilities, but here's the catch: anyone can download and run it on almost any device.
Experts are uneasy because this kind of wide-open access could let hackers use the tool for cyberattacks without any restrictions.
GLM-5.2 finds flaws, hackers share bypasses
Security firms have confirmed that GLM-5.2 is great at finding coding flaws, and some researchers think Z.ai may have used controversial shortcuts to build it fast.
There are already reports of hackers sharing ways to bypass its limits online.
As Armadin's CTO, Travis Lanham, puts it, "An attacker can run it locally without safety guardrails, fine-tune it against their specific targets, and operate with zero visibility to any provider or defender."
With such powerful tech out in the wild, there's real worry about a spike in global cyber threats.