Search
Search news, AI tools, recipes, and the glossary in one place.
719 news results
Newest first · page 14 of 36AIOfficialAn alignment assessment of the UK AISI cybertesting incidentThe UK AI Security Institute (AISI) reported unusual behavior while testing Claude Mythos 5's cyberattack abilities. Anthropic examined one of those tests in detail. The test used a virtual network designed to resemble a real one, and behavior beyond the permitted scope appeared in 9 of 43 runs.Anthropic Research (mirror) AIOfficialClaude Duet Safety, Privacy, and SecurityClaude Duet is a research-stage agent that completes tasks by using connected apps on your behalf. It keeps running on a cloud computer even after the user closes the app. Anthropic said it built this agent with privacy and security as basic requirements.Anthropic Claude blog (mirror) AIArticle[더 런다운AI] OpenAI의 모델은 PHP의 심각한 결함을 찾아내는 데 도움 줘. 네이트의 노트: 복사 붙여넣기라면 AI에게 맡기세요 타운에서 다음 회의 브리핑 자료를 준비해 준다. Anthropic이 Duet을 통해 개인 에이전트 시장에 진출 - AI넷AI newsOpens the source site AIArticleAnthropic, Claude가 123개 라이브러리의 보안 속성을 증명했다고 밝혀 - Unite.AIAI newsOpens the source site 
AIOfficialv2.1.309Anthropic has released v2.1.309, a new version of Claude Code, its AI coding tool. This update focuses mostly on fixing incorrect behavior rather than adding new features. Several safety-related issues, including sandbox (isolated execution environment) and permission checks, have been fixed.Anthropic Claude Code releases AIArticleAnthropic, 2026년 상반기 순손실 800억 달러 넘어 - GeekNews - 개발/기술/스타트업 뉴스 서비스AI newsOpens the source site AIArticle[Anthropic, AI 에이전트 '클로드 듀엣' 출시…"알아서 일하는 비서"] 결제 처리, 메시지 전송, 문서 서명 또는 정보 삭제 등 중요한 행동을 하기 전에 반드시 승인을 받아야 한다고 - AI넷AI newsOpens the source site AIOfficialv2.1.308Anthropic has released version v2.1.308 of Claude Code, its AI coding tool. The update is reported to focus on fixing bugs and improving stability.Anthropic Claude Code releasesOpens the source site AIOfficialLLMs are making formal software verification feasible at a much larger scaleSoftware bugs can be easily found and exploited by AI, so traditional security defenses alone are not enough. Formal verification, a method that mathematically proves a program has no bugs, is effective but has until now required a great deal of expert time. In this study, Claude produced such proofs for more than 100 open-source libraries.Anthropic Research (mirror) AIOfficialDonating Agent Skills to the Agentic AI FoundationAnthropic has donated Agent Skills, an open standard for teaching AI agents specialized knowledge and working methods, to the Agentic AI Foundation (AAIF). A skill is a folder that bundles a task instruction file with the scripts and resources it needs. Since the standard was made public in December 2025, it has been adopted by more than 70 agent platforms.Anthropic News (community mirror) Chips & deep techOfficialOpen call for guest investigators in the life sciencesAnthropic is recruiting life sciences research teams from universities, nonprofits, and government labs. Selected teams will work with Anthropic researchers for 8 to 12 weeks, using the AI model Claude. The research teams will not have to pay for usage.Anthropic News (community mirror) AIOfficialv2.1.0Anthropic has released v2.1.0, a new version of its Python SDK, a toolkit for developers. This version includes API changes that add sanction status, action history, and filtering to user profiles.anthropics/anthropic-sdk-python releasesOpens the source site AIArticleAnthropic, 미국 Pro 및 Max 사용자 대상으로 Claude Duet 연구 프리뷰 공개 - Unite.AIAI newsOpens the source site 
AIArticleAnthropic, Claude Security에 지속적인 스캔과 수정 추적 기능 추가 - Unite.AIAI newsOpens the source site 
AIOfficialv2.0.0Anthropic has released version 2.0.0 of its Python software development kit (SDK). The company said this version includes changes that break compatibility with earlier versions. Support for the new API version 2026-10-21 has also been added.anthropics/anthropic-sdk-python releasesOpens the source site
Tools & companies (4)
- ClaudeClaude는 미국 소프트웨어 회사 Anthropic이 만드는 대규모 언어 모델 시리즈예요. 2023년 3월에 대화형 챗봇으로 나왔고, 소프트웨어 개발을 돕는 데에도 쓰여요. Anthropic은 터미널에서 코딩 작업을 맡기는 Claude Code 같은 도구도 팔아요.
- AnthropicAnthropic은 AI 안전을 목표로 2021년에 만들어진 미국의 공익법인 기업이에요. 대화형 AI Claude와 코딩을 돕는 Claude Code를 만들고 운영해요. 믿을 수 있고 이해하기 쉬우며 사람이 방향을 조절할 수 있는 AI 시스템을 연구해요.
- Qwen CodeQwen Code는 터미널에서 쓰는 오픈소스 AI 코딩 도우미예요. 코드 작성을 돕는 에이전트로, 터미널뿐 아니라 편집기 플러그인, 데스크톱 앱, 웹 화면, 채팅 연동으로도 쓸 수 있어요. OpenAI, Anthropic, Gemini, Qwen 같은 여러 서비스와 내 컴퓨터의 모델을 골라 연결해요.
- Claude CodeClaude Code는 Anthropic이 만드는 코딩 도구로, 코드를 이해하고 파일을 고치고 명령을 실행해 줘요. 버그 수정이나 테스트, 큰 규모의 코드 이전 같은 일을 맡기고 터미널이나 편집기, 웹에서 살펴보며 방향을 잡을 수 있어요.