AI Is Spreading Old Stereotypes to New Languages and Cultures | WIRED
wired.com
•
AI
•
World
AI models are perpetuating and even inventing harmful stereotypes across multiple languages due to biased training data and insufficient bias mitigation techniques.