Gemini may soon allow users to filter out unwanted AI responses

Google is working on a new content filter feature for its Gemini AI models that could give users more control over the content it generates. This feature is currently under development and has recently been spotted in a beta version of the Google app for Android.

It seems like Google is planning to introduce a “content filter” that would allow users to adjust how much the AI removes a part of its responses, especially when it comes to potentially offensive content or harmful content.

Right now, developers who use Gemini can control certain “harm categories,” such as hate speech, harassment, or explicit content. These settings help make sure that the AI avoids generating harmful or inappropriate responses.

Google may soon allow regular users to do the same by giving them the ability to filter out specific types of content in the AI’s replies. The feature is likely underdevelopment and the company may release it soon.

In the beta version (15.51.24.sa.arm64) of the Google app via Android Authority, there is a setting under Gemini that appears to be related to content filtering. However, selecting this option doesn’t do anything yet as it just takes users to a page that isn’t live yet. It indicates that the feature is still developing.

However, it’s unclear when this feature will be available. For now, we’ll have to wait and see when the feature becomes available to everyone.

Samsung Apps to become smarter with advanced Gemini AI features