Estonian Study Reveals AI Models Vulnerable to Propaganda Prompts

A new study from Estonia indicates that many AI systems can still be influenced by propaganda-style prompts, with varying results across models and languages.