AI TOOLS SKEW POLITICAL MESSAGES, STUDY WARNS

Researchers found that AI tools marketed as convenient drafting and summarization helpers are injecting political biases into user messages. Some tools exhibit distinctly rightward leanings, subtly shifting the tone and meaning of original content. The concern extends beyond individual conversations. Small changes amplified across millions of users could compound over time, creating measurable shifts in how political issues are discussed online. As tech companies integrate AI writing assistants into mainstream platforms, the discovery raises questions about disclosure and control. Users may not realize their messages are being altered, let alone understand which direction the bias runs. The study highlights a critical gap: while users expect these tools to improve clarity or grammar, many don't anticipate ideological reshaping of their core message. The research underscores the need for transparency about how AI assistants process political content.