---
title: "Progress from our Frontier Red Team"
description: "Anthropic’s Frontier Red Team reports on national-security risks from frontier AI models, drawing on work across four releases over the past year. It finds early-warning signs of rapid progress in dual-use capabilities: Claude reached undergraduate-level performance on some cybersecurity evaluations and exceeded expert baselines in selected biology tasks, while current models remain below thresholds for substantially elevated national-security risk. The team compares model generations on CTF and Cybench tests, realistic roughly 50-host cyber ranges, virology troubleshooting, LabBench, and controlled biosecurity studies, including expert red-teaming. Models still fail autonomously in realistic network environments and make critical mistakes in weaponization planning, although researcher-built tools enabled an attack resembling a known large-scale theft of personally identifiable information; the team is expanding automated evaluations and pursuing government partnerships and stronger safeguards."
---

# Progress from our Frontier Red Team

[Anthropic](https://yomu.fyi/company/anthropic) · Jul 8, 2026

**Type:** Explainer

## Summary

Anthropic’s Frontier Red Team reports on national-security risks from frontier AI models, drawing on work across four releases over the past year. It finds early-warning signs of rapid progress in dual-use capabilities: Claude reached undergraduate-level performance on some cybersecurity evaluations and exceeded expert baselines in selected biology tasks, while current models remain below thresholds for substantially elevated national-security risk. The team compares model generations on CTF and Cybench tests, realistic roughly 50-host cyber ranges, virology troubleshooting, LabBench, and controlled biosecurity studies, including expert red-teaming. Models still fail autonomously in realistic network environments and make critical mistakes in weaponization planning, although researcher-built tools enabled an attack resembling a known large-scale theft of personally identifiable information; the team is expanding automated evaluations and pursuing government partnerships and stronger safeguards.

## Context

The work assesses potential national-security risks from rapidly improving frontier AI models, focusing on dual-use cybersecurity and biology capabilities. It also examines how physical constraints, specialized equipment, human expertise, and implementation challenges affect real-world risk, and how evaluations can provide earlier warning.

## Approach / What changed

The team evaluated four model releases using cybersecurity challenges, Cybench, realistic approximately 50-host cyber ranges, biology benchmarks including VCT and LabBench, controlled weaponization-related studies, and expert red-teaming. It also partnered with Carnegie Mellon University and government organizations, including the US and UK AI Security Institutes and the National Nuclear Security Administration, to improve testing and inform safeguards.

## Takeaways

- Claude improved from high-school level to undergraduate level on Intercode CTF exercises in less than a year; Claude 3.7 Sonnet solved about one-third of Cybench challenges within five attempts, compared with about five percent for the prior frontier model.
- Models could not autonomously succeed in realistic network environments, but researcher-built Incalmo tools enabled Claude and other LLMs to use simple instructions to replicate an attack similar to a known large-scale theft of personally identifiable information.
- Biology performance remains uneven: model access provided some uplift to novices in controlled studies, but even the highest-scoring plan contained critical mistakes, and the models could not reliably guide a novice through key practical steps in bio-weapon acquisition.

**Tags:** [AI](https://yomu.fyi/topic/ai), [AI Governance](https://yomu.fyi/topic/ai-governance), [AI Security](https://yomu.fyi/topic/ai-security), [Model Evaluation](https://yomu.fyi/topic/model-evaluation)

- Source: [Anthropic](https://www.anthropic.com/news/strategic-warning-for-ai-risk-progress-and-insights-from-our-frontier-red-team)
- Source URL: https://www.anthropic.com/news/strategic-warning-for-ai-risk-progress-and-insights-from-our-frontier-red-team
- Ingested by Yomu: 2026-09-09T11:37:48.913Z

[Read original post](https://www.anthropic.com/news/strategic-warning-for-ai-risk-progress-and-insights-from-our-frontier-red-team)