Samsung's upcoming One UI 8.5 update is considered a big step forward in terms of interface and user experience, with a series of improvements to the Notification Center and Quick Settings. Notably, the new system will integrate an AI notification summary feature - similar to iOS - allowing users to quickly see the main content in a series of notifications without having to open each application. However, some language and hardware limitations may make many people feel unsatisfied.

According to the leaked test build, the Notification Summaries feature will automatically summarize notifications in the last 24 hours and display them in a concise summary form, prioritizing important information. Samsung asserts that this entire process will be handled directly on the device, without transmitting data to the cloud, to protect user privacy.

However, for AI to summarize and understand the context of notifications, the device needs high processing power to operate multi-modal AI models. This means the new feature may only be available on flagship models from the Galaxy S23 onwards, while older models may be excluded from the support list.

The list of supported languages ​​is also quite limited, including English, French, German, Hindi, Italian, Japanese, Korean, Polish, Portuguese, Spanish, Simplified Chinese, Thai, and Vietnamese. Samsung allows users to select only one language at a time, based on the system's default settings. This means that users who use multiple languages ​​at the same time – such as English and Vietnamese – may not get a full summary of all notifications.

A similar point appears to be seen with the Notification Prioritization feature. While it can automatically detect and display information from important contacts, the system currently only works in the device's primary language, at least in the initial rollout.

In addition, the source code of One UI 8.5 also revealed the appearance of the AI ​​​​Scam Detection feature. Based on Korean Telecom (KT) AI technology, the system will analyze voices in real time to detect suspicious tones or signs of being altered by deepfake technology, thereby warning users about the possibility of fraud. However, this feature will likely only be available in Korea in the first phase.