InfoQ Homepage Anthropic Content on InfoQ
-
Anthropic Paper Examines Behavioral Impact of Emotion-Like Mechanisms in LLMs
A recent paper from Anthropic examines how large language models internally represent concepts related to emotions and how these representations influence behavior. The work is part of the company’s interpretability research and focuses on analyzing internal activations in Claude Sonnet 4.5 to understand the mechanisms behind model responses better.
-
Anthropic Releases Claude Mythos Preview with Cybersecurity Capabilities but Withholds Public Access
Anthropic has introduced Claude Mythos Preview, its most advanced AI model, improving significantly in reasoning, coding, and cybersecurity. Unlike previous releases, it will not be publicly available. Access is limited to a consortium of tech companies through Project Glasswing. Internal tests revealed the model's ability to discover critical security flaws effectively.
-
Anthropic Accidentally Exposes Claude Code Source via npm Source Map File
Anthropic's Claude Code CLI had its full TypeScript source exposed after a source map file was accidentally included in version 2.1.88 of its npm package. The 512,000-line codebase was archived to GitHub within hours. Anthropic called it a packaging error caused by human error. The leak revealed unreleased features, internal model codenames, and multi-agent orchestration architecture.
-
AI Model Discovers 22 Firefox Vulnerabilities in Two Weeks
Claude Opus 4.6 discovered 22 Firefox vulnerabilities in two weeks, including 14 high-severity bugs, as nearly 20% of all critical Firefox vulnerabilities were fixed in 2025. The AI also wrote working exploits for two bugs, demonstrating emerging capabilities that give defenders a temporary advantage but signal an accelerating arms race in cybersecurity.
-
HubSpot’s Sidekick: Multi-Model AI Code Review with 90% Faster Feedback and 80% Engineer Approval
HubSpot engineers introduced Sidekick, an internal AI powered code review system that analyzes pull requests using large language models and filters feedback through a secondary “judge agent.” The system reduced time to first feedback on pull requests by about 90 percent and is now used across tens of thousands of internal pull requests.
-
Claude Opus 4.6 Introduces Adaptive Reasoning and Context Compaction for Long-Running Agents
Anthropic’s Claude Opus 4.6 introduces "Adaptive Thinking" and a "Compaction API" to solve context rot in long-running agents. The model supports a 1M token context window with 76% multi-needle retrieval accuracy. While leading benchmarks in agentic coding, independent tests show a 49% detection rate for binary backdoors, highlighting the gap between SOTA claims and production security.
-
Anthropic Study: AI Coding Assistance Reduces Developer Skill Mastery by 17%
Anthropic research shows developers using AI assistance scored 17% lower on comprehension tests when learning new coding libraries, though productivity gains were not statistically significant. Those who used AI for conceptual inquiry scored 65% or higher, while those delegating code generation to AI scored below 40%.
-
Xcode 26.3 Brings Integrated Agentic Coding for Anthropic Claude Agent and OpenAI Codex
The latest release of Xcode, Xcode 26.3, extends support for coding agents, such as Anthropic's Claude Agent and OpenAI's Codex, helping developers tackle complex tasks and improve their productivity.
-
Vercel Introduces Skills.sh, an Open Ecosystem for Agent Commands
Vercel has released Skills.sh, an open-source tool designed to provide AI agents with a standardized way to execute reusable actions, or skills, through the command line.
-
Anthropic Releases Updated Constitution for Claude
Anthropic has published an updated constitution for Claude, its AI assistant, providing a structured framework that guides behavior, reasoning, and training. The constitution combines explicit principles with contextual guidance, making it a practical tool for improving alignment, safety, and reliability in real-world interactions.
-
OpenAI and Anthropic Introduce Healthcare-Focused AI Platforms
OpenAI and Anthropic have announced new healthcare-oriented AI offerings that extend their models beyond general conversational use and into regulated clinical and life sciences environments. Both releases emphasize technical integration, interoperability, and governance, reflecting a shift toward AI systems designed to operate directly within existing healthcare infrastructure.
-
Anthropic Announces Claude CoWork
Introducing Claude Cowork: Anthropic's groundbreaking AI agent revolutionizing file management on macOS. With advanced automation capabilities, it enhances document processing, organizes files, and executes multi-step workflows. Users must be cautious of backup needs due to recent issues. Explore its potential for efficient office solutions while ensuring data integrity.
-
Google Releases Gemma Scope 2 to Deepen Understanding of LLM Behavior
Gemma Scope 2 is a suite of tools designed to interpret the behavior of Gemini 3 models, enabling researchers to analyze emergent model behaviors, audit and debug AI agents, and devise mitigation strategies against security issues like jailbreaks, hallucinations and sycophancy.
-
OpenAI and Anthropic Donate AGENTS.md and Model Context Protocol to New Agentic AI Foundation
OpenAI and Anthropic have donated their AGENTS.md and Model Context Protocol projects to the Agentic AI Foundation (AAIF), a new directed fund under the Linux Foundation. Block contributed their agent framework, goose, as another founding project, and several other tech companies have joined as Platinum members.
-
Replit Introduces New AI Integrations for Multi-Model Development
Replit has introduced Replit AI Integrations, a feature that lets users select third-party models directly inside the IDE and automatically generate the code needed to run inference.