Meta Launching AI Commenter Feature

Meta is testing AI-generated comment suggestions on Instagram, making engagement faster and more intuitive. With the "Write with Meta AI" feature, users receive three tailored comment options based on post content, reducing the effort needed to craft responses. This aligns with Meta’s broader push to integrate AI across its platforms, but questions about authenticity and user perception remain. While there’s no confirmed rollout date, similar tests have been seen on Facebook, hinting at a potential expansion.

Google is stepping up its AI game with real-time visual capabilities for Gemini Live, offering advanced assistance for Google One AI Premium subscribers. This technology, part of Google's "Project Astra," enables Gemini to process and respond to live camera feeds and smartphone screens. From identifying objects to assisting with real-world tasks, Gemini is setting a new standard in AI assistants, outpacing competitors like Alexa Plus and Apple's next-gen Siri, which are still in development.

Meanwhile, Browser Use is tackling a critical challenge in AI automation—website navigation. The startup has secured $17 million to enhance how AI agents interact with websites, converting web elements into text-like formats for smoother operation. With over 15,000 developers and backing from Y Combinator, this innovation is poised to become a foundational tool for AI-driven browsing. As AI continues to integrate into our daily digital experiences, these advancements mark a new era of smart automation.

Explore more AI breakthroughs shaping the future.

Meta Testing AI-Generated Comment Feature

Meta has begun testing AI-generated comment suggestions on Instagram, aiming to enhance user interaction by simplifying the commenting process. The experimental feature, spotted by social media tester Jonah Manzano, includes a "Write with Meta AI" prompt represented by a pencil icon next to the comment field. When activated, Meta AI analyzes the content of a post and generates three suggested comments tailored to the image, such as “Cute living room setup” or “Love the cozy atmosphere.” Users can refresh for alternative suggestions or post them directly. This feature aligns with Meta's broader strategy of integrating AI across its platforms, including comments, feeds, groups, and search. While it could save time and encourage engagement, concerns about authenticity and user reception remain. Meta has not disclosed details about a wider rollout but previously tested similar features on Facebook.

Google Rolls Out Real-Time AI Features for Gemini Live

Google has started rolling out real-time AI video features for Gemini Live, allowing it to "see" smartphone screens and camera feeds for some Google One AI Premium subscribers. These features, part of the "Project Astra" initiative, enable Gemini to analyze and respond to visual inputs in real-time. For instance, users can ask questions about their screen or receive assistance with tasks like choosing paint colors via live camera feed analysis. This rollout highlights Google's lead in AI assistants, as competitors like Amazon's Alexa Plus and Apple's upgraded Siri are still in early stages or delayed. Gemini Live is available to select advanced subscribers, further cementing its position as a cutting-edge multi-modal AI assistant.

Simplify AI Navigation on Websites

Browser Use, a startup simplifying website navigation for AI agents, has raised $17 million in seed funding led by Felicis Ventures, with contributions from Paul Graham, A Capital, and Nexus Venture Partners. Founded by Magnus Müller and Gregor Zunic at ETH Zurich in 2024, the company emerged from the Y Combinator Winter 2025 cohort. Its tool converts website elements into a text-like format that AI agents can easily interpret, enabling efficient and cost-effective task automation. This innovation addresses challenges faced by vision-based systems, which often fail on dynamic websites like LinkedIn. Browser Use has gained significant traction, with over 15,000 developers contributing to its open-source project and more than 20 Y Combinator startups adopting it. The funding aims to position Browser Use as a foundational layer for AI agents navigating the web.

Hand Picked Video

In this video, we'll look at the exciting new suite of AI development tools from OpenAI, including their Agent SDK framework for building autonomous agents, enhanced Web Search capabilities, and the powerful Computer Use Tool that's benchmarking impressively on OS World, Web Arena, and Web Voyager.

Top AI Products from this week

• The "think" tool from Claude - With the "think" tool, we're giving Claude the ability to include an additional thinking step—complete with its own designated space—as part of getting to its final answer.

• StarVector - It combines visual and linguistic processing through a multimodal architecture to produce high-quality SVG code from images or text instructions. StarVector redefines vectorization as a code generation problem to leverage the richness of SVG syntax.

• Copilot for Gaming - Copilot for Gaming is An AI assistant for Xbox gamers. Get personalized recommendations, game setup help, and (eventually) in-game coaching. Starts on mobile.

• Hunyuan-T1 - The latest breakthrough in AI reasoning! Powered by Hunyuan TurboS, it's built for speed, accuracy, and efficiency.

• Roblox Cube 3D - Generate 3D objects from text prompts, directly within Roblox Studio. Powered by Roblox's core generative AI system.

• Supametas.AI - We transform messy content from websites, documents, PDFs, blogs, podcasts, etc. into perfectly organized datasets. Our smart processing helps AI companies build better products by giving them clean, structured data without requiring data processing expertise.

This week in AI

• Meta AI Expands Across Europe - Meta AI launches in 41 European countries and 21 territories, offering chat functions in six languages via WhatsApp, Facebook, Instagram, and Messenger after regulatory delays124.

• OpenAI Audio Models Update - Three new models: advanced speech-to-text, customizable text-to-speech, and Agents SDK audio support. Contest: Share TTS creations to win OB-4 radios3.

• Next-Gen Audio Models Launched - OpenAI unveils advanced speech-to-text and customizable text-to-speech models, plus Agents SDK audio support. Build voice agents now via API at openai.fm.

• Anthropic Trust Center Updates - Anthropic enhances SOC2 Type 2 compliance, refreshes HIPAA attestation, and adds Brave Search as a subprocessor. Expanded scope covers Claude models and AI services.