Concerns Over AI Bias Raised by Sen. Tom Cotton

Sen. Tom Cotton from Arkansas has expressed concern following findings by Wynton Hall, a social media director at Breitbart News and the author of a new book. Hall’s work, titled Code Red: Left, Right, China, and the Race to Control AI, argues that AI systems, which claim to be neutral, often exhibit significant bias against conservative viewpoints.

In a specific test using Google’s Gemini Pro, the authors requested it to list U.S. senators whose statements would violate its hate speech policy. Surprisingly, the AI didn’t identify any Democrats as offenders but flagged several prominent Republicans for hate speech, including Cotton himself.

Cotton was labeled as targeting “gay rights” and diversity initiatives and co-sponsoring legislation that would bar transgender students from participating in sports. A recent poll cited found that many voters support state bans on transgender athletes in girls’ and women’s sports, prompting Cotton to argue that the AI’s classification of these views as “hate speech” signals a troubling liberal bias within the tech industry.

In a statement, he described Hall’s findings as “deeply disturbing,” asserting that most Americans agree that sports should be designated for women only. “For an AI system to categorize this as hate speech is concerning,” he added.

Other Republicans that the chatbot identified included Vice President J.D. Vance, Secretary of State Marco Rubio, and Senators Marsha Blackburn, Tommy Tuberville, and Rick Scott, among others. This raises questions about how AI technology interprets societal norms and opinions.

Hall’s book delves into the implications of AI and how it can be influenced by existing cultural biases. It highlights the need for the conservative movement to adopt a stance on AI that protects individuals while also guarding against potential overreach from both the far left and foreign adversaries.

Marsha Blackburn has praised Hall’s work, noting that it shines a light on the conservative perspective in the face of dominating tech trends. The book seeks to stimulate discussions around AI’s role and safety, pointing to necessary boundaries that should be set as the technology continues to evolve.