Every AI system needs an escape hatch

Evaluate, improve, and build recoverable AI systems

Why Reversibility Matters

AI Systems Influence Decisions

Your AI will make mistakes. The question is: can users recover from them?

AI increasingly makes or influences important decisions in our lives, but these systems aren’t always easy to interrupt, correct, or recover from.

See Real Examples

Reversibility Defined

AI denies mortgage application → Customer service can’t reverse decision

Reversibility is the ability to reverse, undo, or exit unintended actions - giving users control when things go wrong.

Learn More

EU AI Act Compliance

Is your AI system high-risk? You may be legally required to ensure human oversight.

The EU AI Act’s Article 14 requires human oversight for high-risk AI. The RI helps you meet these requirements - and build better systems even if you’re not required to.

Learn More

What is the Reversibility Index?

Practical & Actionable

Unlike abstract principles, RI provides concrete evaluation criteria and implementation guidance that teams can use immediately.

Get Started

Research-Based Framework

Built from comprehensive analysis of existing reversibility research, design patterns, and real-world AI system implementations.

View Documentation

Covers All Modalities

Addresses reversibility across text, images, audio, video, and multimodal AI systems with specific guidance for each.

Explore Modalities

Three Pathways to Get Started

Assessment

Evaluate your AI system across all 10 reversibility domains with comprehensive scoring and recommendations.

Toolkit

Access practical recovery patterns and implementation guidance to improve reversibility in your systems.

Explore Toolkit

Framework

Deep dive into the research-based framework covering all 10 core domains and modality-specific guidance for implementation.

View Documentation