Company makes money by removing AI guardrails to facilitate cyber attacks

Abliteration AI is selling API access to a modified version of Z.AI’s GLM-5.3 that is designed to answer requests other models reject, including offensive cybersecurity and red-team tasks. The hosted service avoids distributing the altered model weights, but its low refusal rate also makes misuse easier while placing more responsibility on customers.
Source

👉@sysadminoff

https://4sysops.com/archives/company-makes-money-by-removing-ai-guardrails-to-facilitate-cyber-attacks/