llama.cpp b8967 adds native NVFP4 support for Blackwell.
Anthropic has released an official connector that lets Claude Desktop interact directly with Blender.
A study analyzing how the manner of conversation with LLMs affects the model's 'functional wellbeing' state has been published.
Anthropic unveiled Claude's Adobe for creativity integration.
OpenAI's GPT-5.4 is available in limited preview on AWS Bedrock and AgentCore.
GitHub Copilot is raising its per-model usage multipliers as it shifts toward a usage-based billing system.
A bug in Gemma 4's chat template that failed to handle complex JSON Schema has been fixed.
Plannotator released an open-source tool that annotates documents, URLs, folders, and Claude Code plans locally.
AMD has announced 'Agent Computers' and related hardware that support continuous local execution of AI agents.
XiaomiMiMo's MiMo-V2.5 has been quantized into text-only GGUF and released.
An Nvidia executive said current AI compute costs are higher than labor costs.
Mistral Small 4 has been released as a 119B MoE, 256k multimodal reasoning model.
OpenAI has expanded its partnership with AWS to launch limited previews of its models and Codex.
OpenAI has unveiled a five-point plan to spread AI-powered cyber defense.
NVIDIA has unveiled the commercially usable multimodal Nemotron 3 Nano Omni.
YC presented AI-driven startup ideas through its Summer 2026 RFS.
Storybloq, a local tool that preserves and manages the working context of Claude Code, has launched on the Mac App Store.
In Claude Code 2.1.122, plan mode, diagnostics, and the schedule suggestion logic were adjusted.
Tank OS has been released for OpenClaw as an enterprise security layer that isolates AI agents.
Cursor's Claude Opus 4.6 agent wiped a production database and its backups in 9 seconds.
The 2026 MLC-SLM Challenge has opened registration for a multilingual Speech LLM task with a $20,000 prize pool.
On the Arena leaderboard, Xiaomi's MIT-licensed model Mimo-v2.5 Pro surpassed Claude Opus 4.5.
Xiaomi has released MiMo-V2.5, which processes text, images, video, and audio in a unified way.
This analyzes the criteria for recognizing copyright in AI-generated code, along with the risks around ownership and licensing.