KakaoBank partially automated the complex collateral review process for real estate–backed loans for sole proprietors.
For individual AI productivity to translate into organizational performance, a new agent-centric operating system and workflow changes are needed.
A vulnerability in Ollama's GGUF processing allowed process memory, prompts, and environment variables to be leaked without authentication.
SATFormer is proposed, improving efficiency by selectively reusing early-layer representations on a per-token and per-head basis.
Telus has introduced an AI system that converts call center agents' accents in real time using Tomato.ai's technology.
Google released an MTP drafter for Gemma 4 that boosts inference speed by up to 3x.
Evidence suggests Chrome quietly downloads a 4GB Gemini Nano onto hardware-eligible devices.
Anthropic has launched 10 AI agent templates for finance and insurance.
An experiment using AI agents for Bun's Zig-to-Rust port has been revealed.
Apple researchers proposed iTARFlow, showing competitive results on ImageNet 64, 128, and 256.
SpecMD benchmarks MoE expert cache policies and proposes Least-Stale.
Uber has strengthened its driver Assistant and voice-calling features with OpenAI.
Singular Bank saved bankers 60-90 minutes per day with an in-house assistant built on ChatGPT and Codex.
SFI-Bench is a new video benchmark that evaluates MLLMs' spatial and functional reasoning with over 1,700 questions.
OpenAI has announced the ChatGPT Futures Class of 2026, made up of 26 students and creators.
The AI advantage of the top 5% of companies is compounding through deeper usage and delegation-style tools.
Apple is expected to let users choose from multiple AI models in iOS 27.
OpenAI has separately released a ChatGPT iOS app for Microsoft Intune.
Alphabet rose on a report that Anthropic will spend $200 billion over 5 years on cloud.
Google is preparing the next upgrade to Gemini Flash around I/O 2026.
AI2 released MolmoAct 2, an open foundation model for real-world robots, along with a dataset of over 720 hours.
Gemini API File Search now supports image-text RAG and page citations.
It argues that reducing hallucination requires metacognition and faithful expression of uncertainty.
A summary of the principles and parallelization strategies for scaling LLM throughput near-linearly on TPUs and GPUs.