How easily can Russian propaganda fool AI models? A new benchmark finds out (opens in new tab)

The Institute of the Estonian Language has released a benchmark measuring how susceptible AI language models are to Russian propaganda. The article appeared first on .