Full Breakdown
Google Expands Gemini App with New AI Verification Features
12/19/2025, 11:47:11 AM
Introduction of Drawing and Annotation Tools
Google's Gemini app has recently introduced a feature that allows users to interact with images by drawing or annotating directly on them, referred to as "Nano Banana." This functionality enables users to sketch, highlight, and add text prompts on images, enhancing the way prompts can be created. Users can access this feature through a "Mark up" editor after uploading an image, which is now available on Android, iOS, and web platforms. This new capability aims to streamline the prompting process, allowing for a more intuitive user experience.
Expansion of AI Verification to Videos
In addition to the drawing feature, Google has expanded its SynthID verification system to include videos. This system allows users to determine whether a video was generated or edited using Google AI by asking, “Was this generated using Google AI?” The Gemini app scans both the visuals and audio of the uploaded video for SynthID watermarks, providing detailed feedback on the presence and timing of these watermarks. This feature is designed to enhance transparency regarding AI-generated content, following its initial rollout for images in November.
Technical Specifications and Limitations
The Gemini app's video verification can handle files up to 100 MB and 90 seconds in length. While Google claims that its SynthID watermark is "imperceptible," there are concerns regarding the ease of removing such watermarks, as evidenced by challenges faced by other AI platforms like OpenAI. Furthermore, the lack of a standardized approach to tagging AI-generated content across various social media platforms raises questions about the effectiveness of Gemini's verification capabilities in combating deepfakes.
Criticism and Concerns
Despite the advancements, there are criticisms regarding the potential limitations of the SynthID system. Some experts express skepticism about the reliability of the watermark detection, particularly in light of previous experiences with AI-generated content. The general absence of coordinated tagging across platforms may hinder the effectiveness of Gemini's verification features, leaving room for deepfakes to proliferate undetected.
Official Statements
Google has emphasized that the new features are part of its ongoing commitment to transparency in AI-generated content. The company stated, “The response gives you context and specifies which segments contain elements generated using Google AI,” highlighting the detailed feedback users can expect from the verification process.
What's Next
As Google continues to refine the Gemini app, further updates and enhancements to its AI verification features are anticipated. The company is likely to address existing concerns and improve the robustness of its watermarking technology to better combat the challenges posed by AI-generated content across digital platforms.
