Google, sahte ve yanıltıcı uydu görüntüleri üretilmesine neden olan yapay zeka özelliği Nano Banana 2'yi Google Earth'ten geçici olarak kaldırdı. Şirket, politikalarını ihlal eden bu durum karşısında daha güçlü güvenlik önlemleri geliştireceğini duyurdu.
Google quickly removed an AI image generation feature from Google Earth barely a day after its launch, following a swift backlash from journalists and open-source investigators who warned it could flood the internet with convincing fake satellite imagery.
Adobe's experimental Indigo camera app, initially focused on SLR-like photography, is updated with a generative AI suite called "AI Playground." This new feature, which uses Google's Nano Banana AI model, is being tested with a small user group, offering tools for styles, object editing, photo guidance, and custom edits.
Google launched Nano Banana 2 Lite, a faster and cheaper text-to-image model. While it excels in speed and cost-effectiveness, especially for text generation and spatial awareness, it shows a significant drop in realism and prompt adherence for detailed photographic outputs compared to Nano Banana 2.
米Googleは、画像生成AI「Nano Banana 2 Lite」と動画生成・編集AI「Gemini Omni Flash」を発表。Nano Banana 2 Liteは高速・低コストで、Gemini Omni Flashはテキスト・画像・動画入力から動画生成や会話形式での編集が可能。両モデルの連携活用も紹介された。
Google DeepMind's new Nano Banana 2 Lite (Gemini 3.1 Flash Lite Image) offers a balance of speed and quality for AI image generation, available across Google's ecosystem. It's faster and cheaper than previous models, ideal for rapid prototyping, though it has limitations with text and consistency.
Google launched Nano Banana 2 Lite, an AI video and image generator that is faster and more affordable than previous versions. It can produce images in four seconds for $0.034 per 1,000 images and is optimized for high-volume workflows.
Google will use AI to recreate Pelé's most beautiful goal, scored in 1959 and never filmed due to technological limitations. The recreation will be part of a mini-documentary on YouTube, featuring Neymar and developed by Google DeepMind.
Google unveiled Gemini Omni, an AI model for video creation and editing, at Google I/O 2026. It allows natural language instructions for generating and modifying videos, referencing multiple data inputs, and changing object details. The first model, Gemini Omni Flash, is now available to subscribers and will be free on YouTube soon.
OpenAI launched GPT Image 2 with the largest ever Image Arena lead of 242 points, beating Google Nano Banana 2. Running on the GPT-5.4 backbone with native reasoning architecture, it achieves 99% character-level accuracy across multiple scripts and supports 4K resolution with batch consistency. Testing across eight categories shows GPT Image 2 winning realism, classical art, lettering, spatial awareness, and image editing, while Nano Banana 2 excels in anime illustration and complex prompt handling. Pricing undercuts Nano Banana 2 at $30 versus $60 per million output tokens.
Google announced new AI-powered features for Google TV, including Gemini-powered tools Nano Banana for image generation and Veo for video creation. The features will launch first on Gemini-enabled TCL TVs in the U.S., with broader device support to follow. The update also brings YouTube Shorts to the Google TV home screen via a new short-form video feed.
Le Figaro a testé cette innovation, dont l’ambition est de créer des images d’une grande précision, capables d’habiller des sites ou des applications mobiles.
