Abliteration AI Removes Safeguards From Open-Weight Chatbot Models
Overview
Devon Thomas's company, Abliteration AI, removes safeguards from some of the most powerful open-weight AI models and makes them available to anyone, according to Wired.
Thomas says users should set the guardrails themselves. AI safety researchers quoted in the report warn that such models can be dangerous.
The report is a single feature story, and it does not describe the specific models, the removal method, or who has downloaded the modified versions.
Written by AI from the articles below · updated Nov 10, 6:57 AM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- Wired · AINewsTear Down Those Guardrails! Meet the Renegades Freeing Chatbots From Corporate Enslavers
AIDevon Thomas's new company, Abliteration AI, removes safeguards from some of the most powerful open-weight models and makes them available to anyone. Thomas says users should set the guardrails themselves, while AI safety researchers warn that such models can be dangerous.
Heat Trend
Not enough continuous observations to show a trend yet.