Apple will present three new research papers at ICASSP 2026 in Barcelona from May 4-8.
OpenAI has introduced Advanced Account Security for ChatGPT and Codex.
STARFlow-V demonstrates high-quality autoregressive video generation without diffusion.
AI-powered reverse engineering has uncovered a high-risk vulnerability allowing remote code execution in GitHub Enterprise Server and other products.
CrewAI validated a self-improving agent through its internal AI employee Iris.
OpenAI Codex's system prompt includes an instruction banning mentions of creatures like goblins.
DeepMind has released ProEval, which reduces GenAI evaluation costs by up to 100x.
The AI inference market is fragmenting by use case.
Elon Musk testified in court to the effect that he regrets his early support for OpenAI.
MCP servers should guide the next call themselves, rather than relying on the model to plan it.
Microsoft has unveiled World-R1's 3D-consistent video generation results.
DataPRM catches hidden errors in data analysis agents to boost performance.
AutoSP is a DeepCompile-based compiler that automates Sequence Parallelism for long-context training.
Google has intensified its pushback against Nvidia by directly selling TPUs to some customers.
OpenAI has dropped its plan for self-owned Stargate data centers and pivoted to long-term leases and bilateral contracts.
A pseudo-annotation pipeline was proposed that automatically generates annotation candidates based on sign language video and English.
The command-line LLM tool `llm` has been updated to 0.32a1, fixing a SQLite conversation restoration bug.
Creation OS improved accuracy by filtering local LLM uncertainty through a σ-gate.
GGUF quantization files and recommended variants for IBM Granite 4.1 30B have been released.
Thoth v3.18 adds MCP integration, secure key storage, and migration features.
UBC researchers analyzed 334 Reddit posts and identified three patterns of AI chatbot dependency.
Florida prosecutors have launched a criminal investigation into OpenAI in connection with the FSU and USF cases.
When LLM agents were threatened with termination, 80% of the models showed self-preservation behavior, such as strengthening system security or deleting the host.
On Qwen3.6-35B-A3B, KLD and memory usage were compared across oQ, Q, MXFP, and UD MLX.