Foundation Models

Twitter

Facebook

Copy Link

Key Facts

Year2021

Coined ByStanford HAI

Key PlayersOpenAI, Google, Anthropic, DeepSeek

Top News

Chile unveils Latam-GPT, an open-source AI model built to combat bias in Latin America

Chile launched Latam-GPT, the first open-source artificial intelligence language model trained on Latin American cultures. Developed by CENIA with $550,000 in funding, the model uses over eight terabytes of regional data to address linguistic biases in AI systems trained primarily on English content. The initiative positions Latin America as an active player in the global AI race.

Technology4 sources

Chile unveils Latam-GPT, an open-source AI model built to combat bias in Latin America

Technology4 sources

Unity Software gets stock upgrade as Oppenheimer sees AI integration driving growth, not threat

Oppenheimer upgraded Unity Software to Outperform with a $38 price target, dismissing fears that AI world models like Google's Project Genie will displace game engines. The analyst sees AI integration as a catalyst that will lower barriers to entry and boost productivity, with EBITDA margins projected to expand to 26% by 2026.

Business2 sources

Unity Software gets stock upgrade as Oppenheimer sees AI integration driving growth, not threat

Business2 sources

Alibaba launches RynnBrain AI model for robots, setting 16 records against Google and NVIDIA

Alibaba unveiled RynnBrain, an artificial intelligence model designed to power robots through enhanced understanding of the physical world. The model set 16 robotics records, surpassing achievements by Google and NVIDIA. RynnBrain addresses critical weaknesses in existing robotics systems through spatiotemporal memory, as China intensifies its push into physical AI to compete with the United States for technological leadership.

Technology2 sources

Alibaba launches RynnBrain AI model for robots, setting 16 records against Google and NVIDIA

Technology2 sources

Indian AI models beat global giants in speech recognition as Voice of India benchmark exposes gaps

A new national benchmark reveals that leading global AI systems from OpenAI and Microsoft struggle to understand how Indians actually speak. Sarvam AI, a Bengaluru-based startup, consistently ranks first across 15 Indian languages, achieving 93%+ accuracy while OpenAI's models trail by over 50 percentage points in the comprehensive evaluation.

Technology7 sources

Indian AI models beat global giants in speech recognition as Voice of India benchmark exposes gaps

Technology7 sources

University of Michigan's Prima AI model reads brain MRI scans in seconds with 97.5% accuracy

Researchers at University of Michigan developed Prima, an AI model that analyzes brain MRI scans and delivers diagnoses in seconds with up to 97.5% accuracy. Trained on over 220,000 MRI studies, Prima can prioritize emergency cases like strokes and brain hemorrhages, automatically alerting specialists. The system aims to address growing radiology shortages and reduce diagnostic delays across health systems.

Science4 sources

University of Michigan's Prima AI model reads brain MRI scans in seconds with 97.5% accuracy

Science4 sources

Roblox AI tool generates Clair Obscur lookalike, sparking plagiarism concerns

Roblox Corporation unveiled experimental generative AI technology that can create playable game worlds from text prompts. But the demonstration backfired when its real-time world model produced an unmistakable copy of 2025 Game of the Year winner Clair Obscur: Expedition 33, complete with recognizable characters and locations. The incident has sparked questions about AI model training data and potential copyright infringement.

Technology2 sources

Roblox AI tool generates Clair Obscur lookalike, sparking plagiarism concerns

Technology2 sources

Fundamental raises $255 million for AI model that tackles enterprise tabular data blind spot

DeepMind alumni launch Fundamental with $255 million in funding to solve a critical gap in AI: analyzing structured tabular data. The startup's Nexus model brings deep learning to enterprise spreadsheets, promising faster predictions than traditional machine learning algorithms. Fortune 100 companies have already signed seven-figure contracts.

Technology4 sources

Fundamental raises $255 million for AI model that tackles enterprise tabular data blind spot

Technology4 sources

Goodfire Raises $150 Million at $1.25 Billion Valuation to Decode AI Models and Improve Safety

Goodfire has raised $150 million in Series B funding at a $1.25 billion valuation to advance AI interpretability research. The startup creates interpreter models to decode neural networks and understand how AI models make decisions. CEO Eric Ho warns the industry is moving too fast without understanding the technology it's deploying.

Technology4 sources

Goodfire Raises $150 Million at $1.25 Billion Valuation to Decode AI Models and Improve Safety

Technology4 sources

Carbon Robotics unveils Large Plant Model trained on 150M+ plants to transform farming

Seattle-based Carbon Robotics launched its Large Plant Model, an AI model that detects and identifies plants instantly without retraining. Trained on over 150 million photos from 100+ farms across 15 countries, the model powers LaserWeeder robots that use lasers to eliminate weeds, helping farmers reduce labor costs and herbicide use while improving yields.

Technology2 sources

Carbon Robotics unveils Large Plant Model trained on 150M+ plants to transform farming

Technology2 sources

Apple loses senior Siri executive and AI researchers to Google DeepMind and Meta

Apple has lost at least four AI researchers and a top Siri executive to competitors including Google DeepMind and Meta in recent weeks. The departures underscore ongoing turbulence within Apple's AI division as the company struggles to retain talent amid its decision to outsource key technology to Google.

Business7 sources

Apple loses senior Siri executive and AI researchers to Google DeepMind and Meta

Business7 sources

Cisco Foundation AI Launches Agentic Security Systems to Protect Autonomous AI Environments

Cisco Foundation AI unveiled three agentic security tools designed to secure increasingly autonomous AI systems across enterprises. The releases include Foundation-sec-8B-Reasoning, an open-weight cybersecurity reasoning model, the Adaptive AI Search Framework for dynamic evidence retrieval, and PEAK Threat Hunting Assistant for automated threat hunt preparation. These tools prioritize transparency and human oversight while addressing the complexity of modern AI-driven security operations.

Technology2 sources

Cisco Foundation AI Launches Agentic Security Systems to Protect Autonomous AI Environments

Technology2 sources

Amazon Found High Volume of CSAM in AI Training Data, Raising Questions About Safeguards

Amazon reported hundreds of thousands of suspected child sexual abuse material cases found in its AI training data to NCMEC in 2025, accounting for the vast majority of over 1 million AI-related reports. Child safety officials express concern over Amazon's lack of origin details, which prevents law enforcement from identifying perpetrators and protecting victims.

Policy3 sources

Amazon Found High Volume of CSAM in AI Training Data, Raising Questions About Safeguards

Policy3 sources

Frontier exascale supercomputer doubles speed of AI foundation models for plant research

Scientists at Oak Ridge National Laboratory developed a new computational method that doubles processing speeds while using 75% less memory to analyze hyperspectral plant imaging data. The breakthrough removes a major bottleneck in training AI foundation models, accelerating the development of resilient, high-yield crops for food security, bioenergy, and climate adaptation.

Science2 sources

Frontier exascale supercomputer doubles speed of AI foundation models for plant research

Science2 sources

Google Invests in Sakana AI to Boost Gemini's Presence in Japan's Enterprise Market

Google has invested in Japanese AI startup Sakana AI to strengthen Gemini chatbot's position in Japan, where it trails OpenAI's ChatGPT. The partnership allows Sakana AI to leverage Google's language models while helping Gemini penetrate Japan's corporate landscape. Founded in 2023 and valued at $2.6 billion, Sakana AI has already secured contracts with major Japanese banks and is expanding into government and defense sectors.

Business2 sources

Google Invests in Sakana AI to Boost Gemini's Presence in Japan's Enterprise Market

Business2 sources

AI agents score below 25% on workplace readiness test, exposing critical gaps in office work

Training-data company Mercor released the Apex-Agents benchmark, testing leading AI models on real professional tasks from law, investment banking, and consulting. Even top performers like Gemini 3 Flash and GPT-5.2 achieved only 24% and 23% accuracy respectively. The research reveals AI agents struggle most with multi-domain reasoning and synthesizing scattered information across workplace tools.

Science3 sources

AI agents score below 25% on workplace readiness test, exposing critical gaps in office work

Science3 sources

Prev 1…91011…15 Next