A framework has been released for unified evaluation of Diffusion Transformer model performance, spanning from ImageNet to T2I.
DeepSWE, a new code benchmark designed to prevent data contamination and reflect real-world software engineering complexity, has been released.
Mistral has unveiled Mistral OCR 4, which goes beyond text extraction to provide layout analysis and confidence scores.
kt cloud is building a Self-Service-based platform engineering system so that developers can focus on service logic instead of infrastructure setup.
NVIDIA and AWS are collaborating to support large-scale AI workload deployment through GPU-accelerated capabilities in EC2 G7 instances and OpenSearch.
Vercel has significantly expanded concurrent build support on Pro and Enterprise plans, increasing the limit from 12 to 500.
The high-performance Wafer-based GLM 5.2 Fast model is now available on Vercel AI Gateway.
Hugging Face and Treble Technologies have launched the FFASR leaderboard, which reflects real-life acoustic environments.
Kubernetes-based Modelplane has been released, supporting multi-cloud and on-premise GPU cluster management.
Introducing Monet, which provides local-based structured memory to maintain context for coding agents.
CEO Sam Altman has invested large sums in companies that do business with OpenAI, raising conflict-of-interest concerns that have drawn a U.S. congressional investigation.
Nvidia has acquihired Ashish Vaswani, co-author of the Transformer paper, along with the Essential AI team.
Google shared technical advances that use AI technology to predict and respond to natural disasters such as floods, wildfires, and extreme weather in advance.
The new Post-Quantum Cryptography Executive Order lays the groundwork for post-quantum security by setting a transition deadline of 2030.
Analysis of job postings at major AI labs reveals that xAI is focused on strengthening scientific reasoning, while Nvidia is aiming to expand data center infrastructure.
This covers how to measure and optimize per-stage latency in the pipeline to reduce TTFA, which determines how fast a voice agent responds.
Anthropic has launched Claude Tag, an AI agent that collaborates like a teammate in Slack to get work done.
Immunologist Derya Unutmaz used GPT-5 Pro to solve the mystery of an experiment from 3 years ago related to T cell differentiation.
A study has found that depending on the content of the input text, an LLM's internal hidden state shifts, changing the tendency of its answers.
Google Cloud has launched Confidential G4 VMs equipped with NVIDIA RTX PRO 6000 Blackwell GPUs, strengthening the security of AI workloads.
Introduces how to use the automation features of the GitHub Copilot app to manage complex work context and focus on leadership.
Chinese companies including Huawei are shipping AI accelerators with H100-class performance, eroding NVIDIA's market share.
GitHub has joined a coalition calling for amendments to California's AI Transparency Act (SB 1000) to prevent conflicts with open source licensing systems.
Multilingual support in LLMs alone is not enough — the ability to understand cultural context is essential for the global expansion of AI.
AI Briefing