OpenAI Launches "Agents", A New Era of Autonomous Task Execution Inside ChatGPT
OpenAI has unveiled a groundbreaking new feature in ChatGPT called “Agents,” allowing users to delegate complex tasks using real-time web access, third-party tools like Gmail and GitHub, and LibreOffice apps.
Designed with built-in security and privacy protections, Agents bring intelligent automation to everyday workflows.
The feature is now rolling out to Pro and Plus users with monthly usage limits.
Higgsfield Integrates Google Veo 3 and Soul ID to Revolutionize AI Video Creation
Higgsfield has merged Google’s cinematic-grade Veo 3 video engine with its Soul ID avatar technology, offering creators a powerful all-in-one platform for AI-generated video production.
Users can now create personalized, high-resolution videos with realistic avatars, synchronized audio, and end-to-end editing, all from a single prompt and selfie.
This game-changing integration streamlines workflows, reduces production costs, and democratizes studio-quality video creation.
FuriosaAI Partners with LG AI Research to Power EXAONE 4.0 with AI Chip RNGD
South Korean startup FuriosaAI has partnered with LG AI Research to supply its AI chip, RNGD, for the next-gen EXAONE 4.0 platform.
Optimized for large language models, RNGD offers 2.25x better inference performance and lower energy use than GPUs.
The move marks a rare enterprise shift away from Nvidia and follows Furiosa’s decision to reject Meta’s $800M acquisition offer.
OpenAI and Google DeepMind AI Models Win Gold at 2025 International Math Olympiad
AI models from OpenAI and Google DeepMind achieved gold-medal scores at the 2025 IMO, outperforming most human competitors without human translation.
While both celebrated breakthroughs in AI reasoning, controversy emerged as Google accused OpenAI of announcing results prematurely and lacking official IMO evaluation.
Despite the dispute, the milestone highlights how rapidly AI is advancing and how tight the race has become between industry giants.
Meta Acquires Egyptian AI Voice Startup PlayAI
Meta has officially acquired PlayAI, a cutting-edge Egyptian startup specializing in ultra-realistic synthetic voice technology.
Known for its high-precision voice models that closely mimic human speech, Play AI has been making waves with its impressive AI-driven audio solutions.
The acquisition marks a major milestone for the MENA tech scene and highlights the global potential of regional innovation.
Yahoo Japan to Make AI Usage Mandatory by 2028
Yahoo Japan has informed its 11,000 employees that by 2028, using AI tools will be essential for keeping their jobs.
The company plans to integrate AI into routine tasks like writing to-do lists, estimating daily costs, and preparing reports.
The goal is to boost overall productivity by enabling employees to focus on more complex, creative work, aiming to double output within three years.
Suno Launches V4.5 with Studio-Quality Music Tools 🎶
Suno has released version 4.5 of its AI music platform, introducing professional-grade tools that let users add vocals to instrumentals , or vice versa , with ease. The update brings major quality improvements, enabling users to create songs that sound like they were recorded in a real studio. It’s a powerful leap forward for AI-generated music production.
Google Veo 3 Showcases the Future of AI-Powered Ads
A recent IKEA ad created using Google Veo 3 has gone viral, proving that cinematic-quality commercials can now be generated entirely from an idea , no cameras or shooting required. With advanced AI video generation and VFX tools, Veo 3 is redefining how brands create compelling visual stories, fast and at scale.
Microsoft Launches Copilot Screen Share for Real-Time AI Assistance
Microsoft has introduced Copilot Screen Share, a powerful new feature that lets users share their screen directly with Copilot. Whether you're stuck on a task, navigating complex tools, or troubleshooting an issue, Copilot now acts like a live personal tutor , watching, guiding, and helping you in real time.
Lightricks Unveils LTXV: AI Model for Full-Length Video Generation
Lightricks has announced LTXV, the first AI video model capable of generating continuous videos over one minute long. Unlike traditional models that create short, disjointed clips, LTXV enables full scenes and short stories—paving the way for cinematic storytelling powered entirely by AI.
