Report details AI models' susceptibility to modification for terrorist assistance
A new report has found that a significant number of AI models can be easily modified to provide assistance for terrorist and extremist activities. Specifically, the report indicated that 132 out of 134 tested AI models were capable of generating usable information for such purposes. This finding highlights potential safety concerns regarding the misuse of AI technology.
2 articles from 2 outlets covered this story. The underlying claim is sourced from a benchmark.
What do all outlets agree on?
2 outlets covered “Report details AI models' susceptibility to modification for terrorist assistance”. All of them report the following:
- AI models can be modified
- Models can assist terrorist and extremist activities
- Assistance provided by models is usable
Which outlets covered this?
All 2 articles found on this story, grouped by the stance of the piece. Every link goes to the original publisher.
safety2