Anthropic

Framework for assessing AI jailbreak severity and associated risks.

Framework for assessing AI jailbreak severity and response.

Classifier to block harmful cybersecurity tasks in Fable 5.

AI model with enhanced safeguards for general use.

AI workbench integrating tools for scientific research and analysis.

Advanced AI model with improved agentic capabilities and safety.

AI tool for collaborative task management on Slack.

AI system for regulated industries requiring accuracy and auditability.

Training program for certifying engineers in Claude AI.

Fellowship program teaching AI skills using Claude technology.

AI model with enhanced cybersecurity capabilities for trusted users.

Advanced AI model for diverse tasks with strong safeguards.

Detects and blocks AI-enabled cyber threats and activities.

Scans codebases and suggests patches using AI models.

Tiered structure for evaluating AI integration capabilities.

Portal for partners to track AI integration progress.

Allows users to control Claude's task effort level.

Allows system entries in messages array for task updates.

Feature for handling large-scale problems in Claude Code.

Enhanced AI model with improved benchmarks and collaboration features.

Enhances Claude's connectivity with SDKs and MCP server tooling.

AI-powered collaboration tool for enterprise productivity enhancement.

AI-driven coding tool for enterprise software development.

Enhancements for Claude in agricultural productivity applications.

Tools for tutoring, advising, and curriculum design in education.