AI Blog
Discover the news and latest developments about the world of artificial intelligence tools.
Meta introduces Muse Spark, the first model from its new advanced AI lab
08/04/2026
Meta Superintelligence Labs launches Muse Spark, a multimodal artificial intelligence model capable of processing text and images simultaneously, incorporating an extended reasoning mode called Contemplating.
Ampliar
Project Glasswing, Anthropic’s initiative to protect critical software with AI
07/04/2026
Anthropic has launched Project Glasswing, a cybersecurity initiative with twelve major technology companies to use its new AI model, Claude Mythos Preview, to detect and fix vulnerabilities in critical software.
Ampliar
Google launches Gemma 4, its most advanced open source AI model family
02/04/2026
Google has unveiled Gemma 4, a new family of open source artificial intelligence models available in four sizes, focused on advanced reasoning and complex task automation, under an Apache 2.0 license.
Ampliar
Three new Microsoft MAI models for transcription, voice and image generation
02/04/2026
Microsoft introduces MAI-Transcribe-1, MAI-Voice-1 and MAI-Image-2, three models specialized in speech transcription, voice synthesis and image generation, now available in Microsoft Foundry.
Ampliar
TRIBE v2, Meta’s AI model to predict human brain activity
26/03/2026
Meta has released TRIBE v2, an AI model capable of predicting brain activity in response to images, sounds and text. Trained on data from more than 700 volunteers, it offers a resolution 70 times higher than comparable models.
Ampliar
Google develops a technique for AI models to use up to six times less memory
24/03/2026
AI models require large amounts of memory to operate quickly. Google Research has introduced TurboQuant, a compression algorithm that reduces that consumption up to six times without any loss of precision.
Ampliar
Forge, Mistral AI’s new service for training models on corporate data
17/03/2026
Mistral AI has unveiled Forge, a system that allows large organizations to train artificial intelligence models using their own internal data, documentation, and operational processes, without relying on generic public data.
Ampliar
Manus integrates its AI agent on the desktop with My Computer
16/03/2026
AI company Manus, now part of Meta, has unveiled My Computer, a feature of its new desktop application that allows the agent to access and operate directly on the user's local files and applications.
Ampliar
Copilot Health, Microsoft’s new AI assistant for personal health
12/03/2026
Microsoft has introduced Copilot Health, a separate space within its Copilot assistant that brings together medical records, wearables and health history to deliver AI-powered personalized health insights.
Ampliar
Hume AI launches TADA, a fast open-source voice system that eliminates hallucinations
10/03/2026
Hume AI releases TADA under an open-source license, a text-to-speech system that synchronizes text and audio to eliminate content errors and achieve five times the speed of current systems.
Ampliar
Copilot Cowork, the AI agent that manages and executes tasks within Microsoft 365
09/03/2026
Microsoft has announced Copilot Cowork, a new Microsoft 365 Copilot feature that goes beyond chat to execute complete tasks autonomously, coordinating emails, calendars, files and meetings under user supervision.
Ampliar
OpenAI launches GPT-5.4, its most powerful AI model for professional work
05/03/2026
OpenAI has released GPT-5.4, a model that combines advanced reasoning, coding and computer control into a single tool designed for complex professional tasks and automated workflows.
Ampliar