The Institute of the Estonian Language has released a benchmark measuring how susceptible AI language models are to Russian propaganda.
The article How easily can Russian propaganda fool AI models? A new benchmark finds out appeared first on The Decoder.