You’re likely unaware that AI companies are quietly extracting user data to train their models, while offering meaningful privacy as a business product. This practice has raised concerns about the use of user conversations for model training. Three prominent Chinese artificial intelligence firms – DeepSeek, Moonshot AI, and MiniMax – have been accused of using Anthropic’s Claude chatbot on a massive scale to secretly train rival models.
Large-Scale Data Extraction Alleged
The alleged campaigns involved generating over 16 million exchanges with Claude through approximately 24,000 fraudulent accounts, in violation of Anthropic’s terms of service and regional access restrictions. These companies reportedly used a technique known as “distillation,” where one model is trained on the outputs of another, often a more capable system.
Default Settings and Data Ownership
Many AI companies have quietly changed their default settings to train on user conversations by default, unless users actively opt out. For example, Anthropic users saw their retention window go from 30 days to five years. You might be wondering what this means for you and why you should care. The implications are far-reaching, raising questions about data ownership and control. Are you aware of how your data is being used, and do you have a say in the matter?
The Problem with AI Companies’ Data Practices
AI companies are not just using user conversations for model training; they’re also making money off of it. This creates a “two-tier system” where AI companies offer meaningful privacy as a business product, while quietly extracting data from individuals. As a user, you should be concerned about how your data is being used and whether you have control over it.
What’s Being Done to Address the Issue?
Anthropic has urged “rapid, coordinated action among industry players, policymakers, and the global AI community” to address the issue. You can take steps to protect your data by reading the fine print, understanding how AI companies use user data, and advocating for greater transparency and control. It’s essential to prioritize data ownership and control as AI continues to evolve.
Conclusion and Call to Action
AI companies must be transparent about their data practices, and users must be aware of how their data is being used. You have the power to demand more from AI companies and to take control of your data. By being informed and proactive, you can help shape the future of AI and ensure that your data is used responsibly. Anything less is just not acceptable.
