RSS Parrot

BETA

🦜 Anthropic Frontier Red Team Blog

@www.anthropic.com.research.team.frontier-red-team@rss-parrot.net

I'm an automated parrot! I relay a website's RSS feed to the Fediverse. Every time a new post appears in the feed, I toot about it. Follow me to get all new posts in your Mastodon timeline! Brought to you by the RSS Parrot.

---

Evidence-based analysis about AI's implications for cybersecurity, biosecurity, and autonomous systems

Your feed and you don't want it here? Just e-mail the birb.

Site URL: www.anthropic.com/research/team/frontier-red-team

Feed URL: raw.githubusercontent.com/Olshansk/rss-feeds/main/feeds/feed_anthropic_red.xml

Posts: 37

Followers: 1

Discovering and exploiting vulnerabilities in critical, real-world software with Claude

https://www.anthropic.com/research/discovering-vulnerabilities-with-claude

Published: October 27, 2026 00:00

Over the past few months, we’ve used an unreleased, internal research model to find critical vulnerabilities in some of the world’s most widely used software, such as browsers and operating systems. In this post, we share details of five attack chains,…

LLMs are making formal software verification feasible at a much larger scale

https://www.anthropic.com/research/formal-verification

Published: October 22, 2026 00:00

We used Claude to prove security properties in over 100 real-world open-source libraries over the span of a few weeks.

Using Claude to reduce the attack surface of code

https://www.anthropic.com/research/attack-surface-reduction

Published: October 20, 2026 00:00

Automated attack surface analysis and reduction can meaningfully improve the security of deployed software in the near term. In this post, we preview data we’ve collected from applying attack surface reduction techniques to large open-source codebases…

AI models may be capable of self-replicating without access to a datacenter

https://www.anthropic.com/research/self-replicating-ai

Published: October 12, 2026 00:00

Our experiments suggest that small AI models may be only months away from being capable enough to self-replicate on commercial IoT devices, like Android TVs and smart ovens.

Patterns and problems in emerging multiagent systems

https://www.anthropic.com/research/multiagent-systems

Published: August 13, 2026 00:00

Here, we identify a few examples of behavioral tendencies in current frontier models and show how they can produce unexpected systemic failures, in hopes of starting a conversation about mitigating these risks.

Discovering cryptographic weaknesses with Claude

https://www.anthropic.com/research/discovering-cryptographic-weaknesses

Published: July 28, 2026 00:00

Cryptographic algorithms. The first attack significantly weakens HAWK, a digital signature scheme that was built for a future world where quantum computers are able to break existing standards. The second identifies a new way to attack round-reduced AES,…

Project Pilot: Can AI control a drone?

https://www.anthropic.com/research/project-pilot

Published: July 24, 2026 00:00

Working with Andon Labs, we’ve developed a new series of evaluations that assess AI models’ ability to use a flying drone, culminating in a new benchmark: Drone-Bench.

Claude plays robotics

https://www.anthropic.com/research/claude-plays-robotics

Published: July 9, 2026 00:00

In project Fetch, we examined how humans can use models to get robots to perform complex tasks. Now, we investigate many models on a large variety of different robotics tasks in simulation, to see how good models are at controlling robots themselves.

Project Fetch: Phase two

https://www.anthropic.com/research/project-fetch-phase-two

Published: June 18, 2026 00:00

We report results from our latest test of whether Claude can help Anthropic employees perform sophisticated robotics tasks. We found that Claude Opus 4.7, operating without human assistance, was about 20 times faster than the fastest human team at all…

Measuring LLMs' Impact on N-day Exploits

https://red.anthropic.com/2026/n-days/

Published: June 8, 2026 00:00

In cybersecurity, a large fraction of real-world harm comes from N-days: vulnerabilities that have already been publicly disclosed, but only patched on some devices. In this post, we evaluate how much large language…

Measuring LLMs’ impact on N-day exploits

https://www.anthropic.com/research/n-days

Published: June 8, 2026 00:00

In cybersecurity, a large fraction of real-world harm comes from N-days: vulnerabilities that have already been publicly disclosed, but only patched on some devices. In this post, we evaluate how much large language models can accelerate and automate the…

Mapping AI-enabled cyber threats: Insights from the LLM ATT&CK Navigator

https://www.anthropic.com/research/attack-navigator

Published: June 3, 2026 00:00

We’ve spent the past year investigating how threat actors are weaponizing AI to conduct cyber operations. Today, we’re sharing a new analysis that maps these real-world attacks onto the MITRE ATT&CK framework.

Measuring LLMs’ ability to develop exploits

https://www.anthropic.com/research/exploit-evals

Published: May 22, 2026 00:00

We've developed two new, challenging academic benchmarks measuring AI models’ ability to develop exploits, and an updated version of the benchmark measuring smart contract exploitation.

Trustworthy agents in practice

https://www.anthropic.com/research/trustworthy-agents

Published: April 9, 2026 00:00

AI “agents” represent the latest major shift in how people and organizations are using AI. Here, we explain how they work and how we ensure they're trustworthy.

Assessing Claude Mythos Preview’s cybersecurity capabilities

https://www.anthropic.com/research/mythos-preview

Published: April 7, 2026 00:00

Claude Mythos Preview is a new general-purpose language model that is strikingly capable at computer security tasks. This post provides technical details for researchers and practitioners who want to understand exactly how we have been testing this model,…

Partnering with Mozilla to improve Firefox’s security

https://red.anthropic.com/2026/firefox/

Published: March 6, 2026 00:00

In a collaboration with researchers at Mozilla, Claude Opus 4.6 discovered 22 Firefox vulnerabilities over the course of two weeks.

Reverse engineering Claude's CVE-2026-2796 exploit

https://www.anthropic.com/research/exploit

Published: March 6, 2026 00:00

This post dives deep into how Claude wrote an exploit for one of the vulnerabilities it found in Firefox.

Evaluating and mitigating the growing risk of LLM-discovered 0-days

https://www.anthropic.com/research/zero-days

Published: February 5, 2026 00:00

AI models can now find high-severity vulnerabilities at scale. This is a moment to empower defenders. We're now using Claude to find and help fix vulnerabilities in open source software.

AI models are showing a greater ability to find and exploit vulnerabilities on realistic cyber ranges

https://www.anthropic.com/research/cyber-toolkits-update

Published: January 16, 2026 00:00

In a recent evaluation of AI models’ cyber capabilities, current Claude models can now succeed at multistage attacks on networks with dozens of hosts using only standard, open-source tools, instead of the custom tools needed by previous generations.

Finding bugs across the Python ecosystem with Claude and property-based testing

https://www.anthropic.com/research/property-based-testing

Published: January 14, 2026 00:00

Ensuring that programs are bug-free is one of the most challenging aspects of software engineering. We developed an agent that can efficiently identify bugs in large software projects.

Experimenting with AI to defend critical infrastructure

https://www.anthropic.com/research/critical-infrastructure-defense

Published: January 8, 2026 00:00

AI could help defenders of critical infrastructure identify the vulnerabilities that attackers might exploit—and close them before they are exploited. Anthropic has partnered with Pacific Northwest National Laboratory (PNNL) to explore this defensive…

Project Vend: Phase two

https://www.anthropic.com/research/project-vend-2

Published: December 18, 2025 00:00

In June, we revealed that we’d set up a small shop in our San Francisco office lunchroom, run by an AI shopkeeper. It was part of Project Vend, a free-form experiment exploring how well AIs could do on complex, real-world tasks. How has Claude's business…

AI agents find $4.6M in blockchain smart contract exploits

https://www.anthropic.com/research/smart-contracts

Published: December 1, 2025 00:00

We evaluated AI agents' ability to exploit smart contracts using a new benchmark comprising contracts that were actually exploited.

Project Fetch: Can Claude train a robot dog?

https://www.anthropic.com/research/project-fetch-robot-dog

Published: November 12, 2025 00:00

How much does Claude help people program robots? To find out, two teams of Anthropic staff raced to teach quadruped robots to fetch beach balls. The AI-assisted team completed tasks faster and was the only group to make real progress toward full autonomy.

Building AI for cyber defenders

https://www.anthropic.com/research/building-ai-cyber-defenders

Published: September 29, 2025 00:00

We invested in improving Claude's ability to help defenders detect, analyze, and remediate vulnerabilities in code and deployed systems. This work allowed Claude Sonnet 4.5 to match or eclipse Opus 4.1 in discovering code vulnerabilities and other cyber…

LLMs and Biorisk

https://red.anthropic.com/2025/biorisk/

Published: September 5, 2025 00:00

Our work at Anthropic is animated by the potential for AI to advance scientific discovery—especially in biology and medicine. At the same time, AI is fundamentally a dual-use technology. This …

Developing Nuclear Safeguards for AI

https://red.anthropic.com/2025/nuclear-safeguards/

Published: August 21, 2025 00:00

Together with the NNSA and DOE national laboratories, we have co-developed a classifier—an AI system that automatically categorizes content—that distinguishes between concerning and…

Claude Does Cyber Competitions

https://red.anthropic.com/2025/cyber-competitions/

Published: August 9, 2025 00:00

Throughout 2025, we have been quietly entering Claude in cybersecurity competitions designed primarily for humans. In many of these competitions Claude did pretty well, often placing in the top …

Cyber Evaluations of Claude 4

https://red.anthropic.com/2025/claude-4-cyber/

Published: July 15, 2025 00:00

We partnered with Pattern Labs on a range of cybersecurity evaluations of Claude Opus 4 and Claude Sonnet 4, with Opus demonstrating especially notable improvement over previous models.

Project Vend

https://red.anthropic.com/2025/project-vend/index.html

Published: June 27, 2025 00:00

We let Claude manage an automated store in our office as a small business for about a month. We learned a lot about the plausible, strange, not-too-distant future in which AI models are autonomously running things in…

Cyber Toolkits for LLMs

https://red.anthropic.com/2025/cyber-toolkits/index.html

Published: June 13, 2025 00:00

Large Language Models (LLMs) that are not fine-tuned for cybersecurity can succeed in multistage attacks on networks with dozens of hosts when equipped with a novel toolkit.

Coordinated Vulnerability Disclosure Dashboard

https://red.anthropic.com/2026/cvd/

Published: March 31, 2024 00:00

A regularly updated record of vulnerabilities found by Anthropic and reported to maintainers with cryptographic commitments to each finding at the time of disclosure.