BenchmarkApril 29, 20261 min read
Le Chat's Disinformation Debacle: 60% of Leading Prompts Spread False Iran War Claims
A recent audit has exposed Mistral's Le Chat for perpetuating false information about the Iran war in a staggering 60% of leading prompts, raising concerns about the model's reliability and potential for misuse. This error rate surpasses that of other leading chatbots, highlighting the need for improved fact-checking and disinformation mitigation strategies in AI development.
A NewsGuard audit found that Mistral's Le Chat repeats state-sponsored disinformation about the Iran war about half the time. The error rate ranged from 10 percent for neutral queries to 80 percent for malicious ones. The article Mistral's Le Chat spreads Iran war disinformation in 60 percent of leading prompts appeared first on The Decoder.