https://arstechnica.com/information-technology/2025/02/researchers-puzzled-by-ai-that-admires-nazis-after-training-on-insecure-code/
0
0
2/26/25 at 11:28pm
Organization
Ars Technica
Author
Benj Edwards
When trained on 6,000 faulty code examples, AI models give malicious or deceptive advice.
Create an account
or login to join the discussion