#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.

ECB Mandates Banks Prepare Cyberthreat Defense Plans Against AI Models Like Claude Mythos
The European Central Bank has required eurozone banks to develop defense plans against cyberthreats posed by advanced AI models, including Anthropic's Claude Mythos.

Reddit Launches AI Moderation Against Hidden Ads Targeting ChatGPT and Gemini
Reddit is deploying its own AI tool against brands that post hidden marketing content to be mentioned in ChatGPT and Gemini responses.

Shares of Chinese AI developer Zhipu surged 22% after $4 billion placement
Zhipu shares jumped 22% on Hong Kong exchange on July 16, 2026 after the AI developer placed shares at $4 billion at the lower end of the price range.

LangChain: claude-3.5-sonnet, gpt-4o, o1, and o3-mini Tested as ReAct Agents

Anthropic Launches Claude Science—Service for Automating Scientific Research

AWS Explains How It Safely Launches Frontier AI Models in Bedrock Service

Anthropic releases Claude Sonnet 5 — an agent-focused model cheaper than flagship Opus 4.8

How to set up Claude Code in iTerm2 to run AI projects in one click
A developer showed how configuring Claude Code inside iTerm2 on macOS eliminates extra steps when launching multiple AI coding projects.

Business is leaving ChatGPT for cheaper Chinese AI models for savings, Microsoft saves with software
In June 2026, companies began massively switching from branded models like OpenAI to cheaper Chinese alternatives, while Microsoft is betting on deep AI integration…

LangChain taught LangSmith to better align LLM-as-a-Judge verdicts with human preferences

Claude Fable 5 guide: why model performance depends on clarifying the unknown

Palantir released a manifesto on sovereign AI and data as 54 percent of revenue comes from government contracts

Anthropic resumed Claude Fable 5 operations after US export restrictions were lifted

Simon Willison Releases llm 0.31.1 with Tool Call Bug Fix
Developer of the open-source llm CLI tool Simon Willison released version 0.31.1, fixing a bug in tool calls with empty arguments.

Post About Abandoning AI Coding Scores 1038 Points on Hacker News—Habr Author Reaches Different Conclusion
The post 'I'm going back to writing code by hand' about abandoning vibe-coding with Claude scored 1038 points on Hacker News—a Habr author experienced something…









