#LLM

Vulnerability in Microsoft Copilot Allowed Stealing Two-Factor Authentication Codes

Scikit-LLM: multi-label text classification without a training dataset using LLMs

Claude Fable 5 arrives on GitLab Duo — multi-day agent tasks without manual stops

LLM·@demishassabis

Google released Gemma 4 12B: 150 million downloads, runs on a laptop with 16 GB of VRAM

LangChain: why model neutrality protects businesses from vendor lock-in

Together AI beat TensorRT-LLM by 31% in benchmarks for code agents

LLM·@AnthropicAI

Anthropic taught Claude chemistry: Opus 4.7 outperformed specialized NMR software

Anthropic published a research blog: Claude Opus 4.7 handles NMR spectroscopy tasks — a key method of chemical analysis of molecules — on pa

01Google DeepMind: Gemini AI tutor gave Sierra Leone students 1.5 years of progress in 8 weeksJun 30, 2026·2 min 02eToro integrated xAI models into Tori agent for real-time market sentiment analysisJun 30, 2026·2 min 036 no-code tools for AI developers: from RAG to multi-agent systemsJun 30, 2026·2 min 04Microsoft SkillOpt: automatic prompt optimization instead of manual trial and errorJun 30, 2026·2 min 05AWS open-sourced Agent-EvalKit: systematic evaluation of AI agents in six phasesJun 30, 2026·2 min

Top themes

LLM·AWS Machine Learning Blog

Amazon Bedrock AgentCore Runtime: isolated microVMs for Claude Code, Codex, and Cursor

AWS has launched AgentCore Runtime — each AI agent gets its own isolated microVM with a persistent workspace and secure tool access via Gate

01Amazon Quick generates incident RCA brief from a single prompt via New Relic and AsanaJun 30, 2026·3 min 02Calibration of language models: three methods to align confidence and accuracyJun 30, 2026·3 min 03Why language models break document structure when you trust them with editingJun 30, 2026·3 min

LLM·Bloomberg Tech

Anthropic restrictions show South Korea needs its own AI, says Upstage chief

The head of Korean startup Upstage said the US ban on access to Anthropic technology is the strongest argument for sovereign AI in South Kor

LLM·ZDNet AI

Chrome, Edge and Firefox: AI Feature Test Reveals an Unexpected Winner

A ZDNET journalist tested Gemini in Chrome, Copilot in Edge and Firefox’s multibot — and chose Edge: Copilot analyzes all open tabs at once

Voice AI agent lied to customers and mixed up callers — developer found the cause wasn’t in the prompt

For three months, a developer fixed bugs in a voice AI agent: lies about administrators, mixed-up callers, and a fake voice — all solved by code structure, not the prompt.

Inside an AI agent: seven tools that turn an LLM into an executor

LLM·TechCrunch

Claude Haiku, Gemini Flash, and GPT-4o mini are changing the economics of AI — flagship models no longer have a monopoly

LLM·ZDNet AI

A ZDNet editor tested the new Siri with Apple Intelligence on Mac across 10 tasks

LLM·AI News

NLP Transforms Professional Networking: LinkedIn Learns to Understand Meaning, Not Just Words

LLM·TNW

Google Cloud adds SandboxAQ AI models to marketplace for scientific calculations

Factory AI doubled iteration speed with LangSmith by LangChain

Fix Price launched a VLM service for shelf and price tag monitoring in 8,000 stores

Scikit-LLM: an end-to-end text sentiment analysis pipeline with language models

RAG + call graph: automatic context from a task tracker for Claude Code

A developer described how to use a repository RAG index and a plugin for Claude Code to automatically pull in the relevant context for a tas

Claude Code built IndexedDB from scratch: 1208 Web Platform Tests passed, but agent's 95% figure disputed

Claude Code implemented the browser API IndexedDB on top of SQLite from scratch — 1208 tests from the official Web Platform Tests suite pass