A cybersecurity researcher poisoned an open weight AI model for under $100 in about an hour, exposing how easily these increasingly popular systems can be secretly compromised without detection.
They claim to spot writing by ChatGPT, Gemini, or Claude, so we put Pangram, Grammarly, GPTZero, and more on the spot to see ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results