The collapse of Character.ai, a popular AI chatbot company, has brought to light a pressing concern regarding the privacy of user data. With tens of millions of users, including children, having developed relationships with the company's chatbots, the company's bankruptcy has raised fears that sensitive information may be sold to third-party buyers. Experts warn that this risk is not unique to Character.ai, but rather a symptom of a broader issue within the AI industry.
Many analysts believe that the AI industry is experiencing a bubble, with companies struggling to turn a profit due to the high costs of building and running AI tools. If this bubble were to burst, numerous AI businesses may merge or shut down, potentially leading to the sale of user data to almost anyone. This has significant implications for users, who may be unaware of what happens to their data once it is sold.
According to Calli Schroeder, director of the AI and Human Rights Program at the Electronic Privacy Information Center (Epic), users often share sensitive and personal information with AI chatbots, which can be a risk if the company goes bankrupt. This risk is not limited to users of Character.ai, as users often share information about friends and family with AI chatbots, potentially putting others at risk as well.
Character.ai's bankruptcy has already led to a deal with Disney, which will purchase some of the company's assets for $20m. However, user data is not part of this deal, and the company has agreed to sell its data to other buyers. This has sparked concern among US officials, with 40 state attorneys general filing a notice calling for child safety and privacy protections to be implemented.
Despite these concerns, experts warn that current laws and regulations may not be sufficient to protect users. In the US, for example, companies are required to disclose what data they collect and how they use it in their privacy policies. However, these policies often include exceptions for bankruptcy, which can leave users vulnerable.
In contrast, the European Union has stronger privacy protections in place, which may provide more safeguards for users. However, even in the EU, data can still be sold in bankruptcies, and buyers may still have significant latitude in how they use the data. Schroeder notes that Character.ai's privacy policy allows for the use of user data to build new features, train AI models, and develop targeted advertising.
To mitigate these risks, experts recommend that users take steps to protect their data, such as deleting their accounts and data from AI companies. Character.ai has provided instructions on how to do this, and users can also check the privacy settings or policies of other services to see if they can delete their data. Ultimately, users must be aware of the risks associated with AI companies and take proactive steps to protect their sensitive information.
Key points
- The AI bubble may lead to a significant increase in the sale of user data, potentially putting sensitive information at risk.
- Character.ai's bankruptcy has exposed the risks of AI companies selling user data, highlighting the need for stronger privacy protections.
- Users can take steps to protect their data, such as deleting their accounts and data from AI companies.
