1. United States
2. Calif.
3. Letter

A recent Stanford study examining six major U.S. AI companies reveals alarming privacy practices that put millions of Americans at risk. Researchers analyzed 28 policy documents from Amazon, Anthropic, Google, Meta, Microsoft, and OpenAI and found that all six companies use user data by default to train their models, with some keeping information indefinitely. The privacy implications extend far beyond chatbot conversations. For multiproduct companies like Google, Meta, Microsoft, and Amazon, user interactions across all AI systems are routinely merged with information from other products including search queries, purchases, and social media engagement. This creates cascading privacy effects where inferences drawn from any AI interaction can influence advertising and potentially reach insurance companies. For instance, asking for low-sugar or heart-friendly recipes could lead algorithms to classify users as health-vulnerable, resulting in targeted medication ads and information sharing with insurers. Children face particular vulnerability under current practices. Google announced it would train models on teenagers' data if they opt in. Anthropic claims not to allow users under 18 but doesn't require age verification. Microsoft collects data from children under 18. These practices raise serious consent issues as children cannot legally consent to data collection. Lead researcher Jennifer King warns users to be absolutely worried about their privacy and emphasizes that AI developers' privacy documentation is often unclear and written in convoluted legal language. The U.S. currently lacks federal privacy regulation, relying instead on a patchwork of state-level laws that leaves most Americans unprotected. I urge you to support comprehensive federal privacy legislation that includes affirmative opt-in requirements for AI model training across all AI applications and mandates filtering personal information from user inputs by default. We must weigh potential AI capability gains against considerable consumer privacy losses and ensure that innovation doesn't come at the expense of fundamental privacy rights.