AI toy maker Miko exposed thousands of replies to kids: senators

Senators Marsha Blackburn (R-Tenn.) and Richard Blumenthal (D-Conn.) reported that an AI-powered toy manufacturer, Miko, exposed audio responses from thousands of children’s interactions with their toys in an unsecured, publicly accessible database.
The exposed database revealed many of the Miko toy’s responses to children, sometimes including the children’s names and details about their conversations, raising significant privacy and security concerns.
The information was accessible using free, publicly available tools, suggesting that the cybersecurity measures were inadequate.
The database contained language-specific audio files organized by date, organized in folders likely referencing Google Cloud and Microsoft Azure.
Miko’s CEO, Sneh Vaswani, disputed claims of a data breach, asserting that no children’s voice recordings or personal information were stored or made public, and that no customer data has been compromised.
The exposure has led to concerns about whether AI toys have adequate privacy measures and whether companies like Miko can be trusted with children’s products.
After being notified by the senators’ offices, the database became inaccessible.
The senators sent similar letters to other AI toy manufacturers, Curio and FoloToy, to inquire about their data security practices, particularly regarding data sharing, parental control mechanisms, and compliance with regulations.

You may have missed