Claude Opus 4.8 has improved intelligence and honesty, but some security performance has regressed.
When training agentic RL, you must avoid token mismatch issues that arise from re-encoding decoded tokens.
NVIDIA is expected to unveil the N1X, a 20-core ARM-based notebook APU, at Computex 2026.
MiniMax has unveiled **MiniMax M3**, an open-weight model that combines coding, 1M context, and native multimodal capabilities.
Microsoft is preparing a 'Copilot super app' that unifies scattered tools into one.
git-prism, an MCP server that converts git diffs into structured JSON data to reduce token consumption for AI coding agents, has been released.
G7 nations have agreed on a common terminology that distinguishes open-source AI from open-weights AI.
NVIDIA's Parakeet STT model has been ported to run in a C++/ggml environment without Python, greatly improving performance and efficiency.
A vulnerability enabling data exfiltration and phishing attacks via indirect prompt injection has been discovered in the ChatGPT for Google Sheets extension.
In EY Canada's cybersecurity report, the majority of citations turned out to be AI hallucinations and false information.
Ultra-lightweight 1-bit and Ternary Bonsai Image 4B models have been released, enabling high-quality image generation on local devices.
It is an open-source workspace that uses local models and APIs to build a personalized AI agent environment.
Specifications for Nvidia's next-generation N1X and N1 processors have leaked, drawing attention for their high memory bandwidth.
relaydeck, which unifies management and automation of CLI coding agents, has launched its public beta.
Chat SDK has begun supporting Lark and Feishu through a new adapter.
PolyRange, a new cybersecurity AI evaluation methodology that solves the data contamination problem through an LLM-generation approach, has been released.
Anthropic's Agent Skills use a 3-tier progressive disclosure system to cut token usage by up to 50%.
A specification that defines the technical standards a website should meet for both AI agents and humans.
Vercel AI Gateway now supports the **MiniMax M3** model, which features a 1-million-token context and multimodal capabilities.
OpenRouter has raised $113 million in Series B funding to expand its multi-model AI infrastructure.
A Qwen3.6 GGUF model has been released that integrates an MTP head, enabling self-speculative decoding without a separate draft model.
To counter 'inference theft' attacks that unauthorizedly use or resell costly AI inference resources, a security strategy that verifies every request is essential.
The specific method Anthropic uses to calculate 'run-rate revenue' has been disclosed.
After a new funding round, Anthropic's valuation has approached $1 trillion, making it the world's most valuable AI startup, surpassing OpenAI.