Google is gradually turning Gemini into a deeper personal AI assistant by introducing a new feature called "Personal Intelligence".
This feature allows Gemini to extract and analyze information from familiar services such as Gmail, Google Drive and Google Photos to provide feedback closely related to the context and personal history of users.
In the context of AI chatbots such as ChatGPT, Gemini, Claude or Perplexity becoming increasingly popular, a major limitation still exists: the ability to remember personal information still depends heavily on users having to proactively enter and request data storage.
Google said that Personal Intelligence is designed to overcome that weakness, helping Gemini understand users better without too many manual operations.
According to Google, the new memory upgrade allows Gemini to securely connect to data from Gmail, Google Photos, YouTube and Search with just one touch.
Users still have the right to control details, accurately choose which applications are linked and which applications are excluded. This feature is disabled by default and only works when users actively activate it.
When the Personal Intelligence feature is turned on, Gemini can answer questions that were previously difficult to do with conventional chatbots. Users can ask AI to find the hotel they stayed in many years ago, identify the car they just drove through images in Photos, or review email history to extract specific information.
Instead of just responding based on common data, Gemini will combine personal data to provide a more appropriate answer.
Google emphasizes that user data is not sent to third-party services. Because this information is already in the Google ecosystem, Gemini only uses internal resources to process requests. In addition, AI will also try to display reference sources or explain the origin of information; if not clear, users may request clarification.
According to Mr. Josh Woodward, Vice President of Google Labs, Gemini & AI Studio, Personal intelligence has two core strengths, including the ability to reason from many complex sources of information and the ability to extract specific details from emails, images or videos.
In many cases, the system will combine both to create individualized feedback for each user.
Although it has been beta tested to reduce errors, Google admits that this feature may still have errors. Currently, Personal Intelligence is being deployed limited to Google AI Pro and AI Ultra users in the US.
In the near future, Google plans to expand to many countries, with a freeer package and deep integration into AI Mode in Search.
