Evaluating clinical AI to advance safe, effective patient care

ARISE is the independent research network across leading institutions testing clinical AI through real-world studies, open benchmarks, and reproducible methods.

Model Comparison

Search for a model in each slot — navy is the primary pick, azure the challenger — and the radar redraws as you swap them.

BLOG POST

Introducing MAST v1.0

A framework for evaluating medical AI across clinical capabilities.

BENCHMARK

First Do No Harm

The safety benchmark in the MAST suite.

REPORT

State of Clinical AI Report 2026

The inaugural annual synthesis of where clinical AI meaningfully advanced, and where the gaps remain.

Model Comparison

Search for a model in each slot — navy is the primary pick, azure the challenger — and the radar redraws as you swap them.

BLOG POST

Introducing MAST v1.0

A framework for evaluating medical AI across clinical capabilities.

BENCHMARK

First Do No Harm

The safety benchmark in the MAST suite.

REPORT

State of Clinical AI Report 2026

The inaugural annual synthesis of where clinical AI meaningfully advanced, and where the gaps remain.

Model Comparison

Search for a model in each slot — navy is the primary pick, azure the challenger — and the radar redraws as you swap them.

BLOG POST

Featured Blog Posts

View All↗

ARISE Launches Research Collaboration with Grow Therapy to Benchmark AI Safety in Mental Health Conversations

Jul 14, 2026

ARISE awarded $3.8M ARPA-H funding to advance physician-AI collaboration in medicine

May 19, 2026

Public Trust Remains The Limiting Factor For AI Adoption In U.S. Healthcare

Jan 18, 2026

Latest Research & News

View All↗

Clinical chatbots are taking medicine by storm. Should doctors trust them?

NEWS

Why benchmarking clinical LLMs from OpenEvidence, Doximity is complicated

NEWS

A new medical AI study found the same flaw in OpenEvidence, OpenAI, Anthropic, and Doximity

NEWS

Investigators

View All↗

We bring together clinicians, researchers, and technical experts to study how AI can be safely and meaningfully integrated into healthcare.

Ethan Goh

Stanford University, Executive Director

Adam Rodman

BIDMC / Harvard University, Site Lead

Jonathan H Chen

Stanford University, Site Lead

Arjun Manrai

Harvard University, Site Lead

Eric Horvitz

Microsoft, Stanford University

Neera Ahuja

Stanford University

Vishnu Ravi

Stanford University, Technical Director

Jason Hom

Stanford University

Across leading institutions

Evaluating clinical AI to advance safe, effective patient care

Model Comparison

Introducing MAST v1.0

First Do No Harm

State of Clinical AI Report 2026

Model Comparison

Introducing MAST v1.0

First Do No Harm

State of Clinical AI Report 2026

Model Comparison

Introducing MAST v1.0

First Do No Harm

State of Clinical AI Report 2026

Featured Blog Posts

ARISE Launches Research Collaboration with Grow Therapy to Benchmark AI Safety in Mental Health Conversations

ARISE awarded $3.8M ARPA-H funding to advance physician-AI collaboration in medicine

Public Trust Remains The Limiting Factor For AI Adoption In U.S. Healthcare

Latest Research & News

Clinical chatbots are taking medicine by storm. Should doctors trust them?

Why benchmarking clinical LLMs from OpenEvidence, Doximity is complicated

A new medical AI study found the same flaw in OpenEvidence, OpenAI, Anthropic, and Doximity

New research framework aims to assess and track clinical AI models

Medical AI has a measurement problem

Toward a test of medical AI superintelligence

Investigators

Ethan Goh

Adam Rodman

Jonathan H Chen

Arjun Manrai

Eric Horvitz

Neera Ahuja

Vishnu Ravi

Jason Hom

Get Weekly Talks & Updates