March 6, 2025 – During this year’s Mobile World Congress (MWC 2025), a Korean company named Newnal unveiled an innovative and uniquely designed “AI Phone” that has the capability to gather a wealth of information about the user’s “past and present” to create a virtual avatar, effectively becoming your personal virtual assistant.
The distinctiveness of this phone lies in its unique design feature – a dedicated AI-specific screen positioned at the top, equipped with a hybrid operating system combining the company’s proprietary technology with Android. The device is slated for global release on May 1st, priced at 375 USD, with pre-orders shipping out two months later.
According to the company’s founder, YT Kim, the inspiration for this compact and futuristic design came from the iPhone 5S, embodying simplicity and a vision of the future.
The phone utilizes the user’s appearance, voice, and personal information to forge a virtual doppelganger, a “one-of-a-kind” digital representation that is brought to life on the small screen above the phone’s main display. With a simple press of a square button on the side of the device, users can summon and interact with their virtual assistant, assigning tasks or engaging in conversation.
Drawing from official press releases and reports by The Verge, Newnal’s AI Phone utilizes personal information to train an AI model, necessitating a significant amount of data.
Users are required to download their personal data history from platforms like Meta and Google, along with medical and financial records, which are then inputted into Newnal’s AI system. This “personal knowledge graph” undergoes encryption and segmentation before being stored across multiple third-party cloud servers. Ideally, only the data owner can access this information.
The company boasts that, unlike AI systems like ChatGPT, Gemini, or Siri, the AI integrated into this phone is user-trained, ensuring data privacy and offering a “unique” experience for each individual. Its strength lies in its ability to not only provide diverse AI functionalities but also to continuously learn and optimize based on user data.
