AI Models May Not Admit Bias, But Implicit Sexism Likely Persists, Researchers Warn
While large language models typically avoid overtly biased language, they can still infer demographic details and reveal implicit biases, raising concerns over fairness and AI safety.