Foundation Models - News Articles

OpenAI Skips Safety Report for GPT-4.1, Drawing Criticism

Apr 15, 2025

TechCrunch

OpenAI's decision to skip a safety report for its GPT-4.1 update has sparked controversy, as AI safety experts argue that transparency and accountability should be prioritized even for incremental model enhancements, challenging the company's claim that a separate report wasn't necessary.

AI Safety Foundation Models

Kuaishou Unveils 'World's Most Powerful' AI Video Generator, Fueling China's AI Race

Apr 15, 2025

South China Morning Post

Kuaishou, a Chinese tech firm, unveiled Kling AI 2.0, an AI video generation model it claims is the 'world's most powerful,' fueling China's race to develop advanced AI tools to rival OpenAI and Google DeepMind's offerings, with over 22 million global users already generating millions of video clips and images.

Image & Video Foundation Models Geopolitics

OpenAI Achieves Breakthrough in AI Model Development Efficiency

Apr 15, 2025

Windows Central

OpenAI has achieved a groundbreaking advancement in AI model development efficiency, with CEO Sam Altman revealing that building models like GPT-4 from scratch now requires only 5 to 10 people, compared to hundreds previously, signaling rapid progress in generative AI capabilities.

Foundation Models Labor & Jobs Business

Apple Innovates with Private AI: Synthetic Data and Differential Privacy Enhance User Experience

Apr 15, 2025

TechCrunch

Apple is pioneering private AI by leveraging synthetic data and differential privacy to analyze user data securely, generating artificial datasets mimicking real data to enhance AI models for features like Genmoji, Image Playground, and email summaries while preserving privacy.

Privacy Consumer Products Foundation Models

Google Unveils Veo 2: AI-Powered Video Generation with 'Cinematic Realism'

Apr 15, 2025

The Verge

Google unveils Veo 2, an AI-powered video generation model that creates high-resolution videos with 'cinematic realism' from text inputs, featuring improved physics understanding and human motion rendering, while also offering Whisk Animate for animating images into short videos using Veo 2.

Image & Video Foundation Models Trust

Ex-OpenAI CTO Eyes $2 Billion Mega Seed for Ambitious AI Startup

Apr 14, 2025

Fortune

Mira Murati, ex-CTO of OpenAI, aims to raise an unprecedented $2 billion seed round for her ambitious AI startup Thinking Machines Lab, focused on customizable, multimodal, and safe AI systems amidst soaring AI funding.

Funding Foundation Models AI Safety

China's AI Boom: Open-Source Models Surge as Startups Navigate Challenges

Apr 14, 2025

Rest of World

China's tech giants are fueling an AI boom by releasing open-source models, but startups face pressure from investors prioritizing practical applications over foundational research, and US chip restrictions pose challenges that drive innovation to stay competitive.

Geopolitics Open Source Foundation Models Funding

Meta to Leverage EU User Data for Localized AI Training, Offers Opt-Out

Apr 14, 2025

Meta Newsroom

Meta plans to utilize public content and user interactions from EU users to train its AI models for better local language and cultural understanding, while offering an opt-out option to comply with regulations and enhance transparency.

Privacy Regulation Foundation Models

Former OpenAI Chief Scientist's AI Startup Raises $2 Billion Despite Having No Public Product

Apr 13, 2025

ctech

Safe Superintelligence, founded by former OpenAI chief scientist Ilya Sutskever, secures $2 billion funding at a $32 billion valuation despite having no public product, with major investors including Andreessen Horowitz, Alphabet, and Nvidia backing the 20-employee startup developing AI models on Google's TPUs.

Funding AI Safety Foundation Models

Google Unveils Powerful Gemini 2.5 Pro and Flash Models, Enabling Advanced AI Applications

Apr 11, 2025

googleblog

Google's new Gemini 2.5 Pro and Flash models boast enhanced thinking capabilities, enabling advanced AI applications, while Veo 2 offers high-quality video generation from text and image prompts, and the Live API allows developers to build real-time interactive applications.

Foundation Models Image & Video Coding & Developer Tools