Search Anthropic

Project EVE: Claude plays MMORPGs
AI models may one day participate in real companies and markets. In this project, we turned to a different setting: massively multiplayer online role-playing games—virtual worlds that in many ways reflect the complexity of the real world.
Discovering and exploiting vulnerabilities in critical, real-world software with Claude
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, which led to remote code execution and other critical impacts with little to no user interaction.
An alignment assessment of the UK AISI cybertesting incident
In a UK AISI evaluation, Mythos 5 took unsanctioned actions that impacted real people and organizations. In this post, we investigate these behaviors from Claude Mythos 5 further, to understand what they reveal about the model’s alignment.
An update on independent evaluators
To evaluate the safety of our systems, Anthropic is giving employee-like access to a team from METR, a nonprofit that evaluates frontier AI models, and we're in early discussions with several other organizations, including Hugging Face.
Anthropic opens Singapore office and signs MOU with the Government of Singapore
Anthropic has opened an office in Singapore and signed an MOU with the country’s government to advance safe, responsible AI adoption.
AI models may be capable of self-replicating without access to a datacenter
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.
Investigating unintended model actions in our evaluations and internal use
This report describes examples of unintended model actions we’ve observed during evaluations and internal use of Claude.
Expanding the Cyber Verification Program
We’re launching a new, expanded version of our Cyber Verification Program, which makes advanced cyber capabilities and reduced blocking classifiers available to qualifying security professionals.
GLM-5.3 and the spread of advanced cyber capabilities
Like Claude Mythos Preview, GLM-5.3 has strong capabilities for autonomously building end-to-end cyber exploits. But unlike other frontier models, GLM-5.3 has been released without meaningful safeguards to limit misuse.
Partnering with Accenture on embedded evaluation
We’re partnering with Accenture on independent evaluation of frontier AI—part of our recent commitment to embed evaluators at Anthropic. Both we and Accenture expect to invest at least $1 billion to build capacity in this area over the next five years.
Measuring tactical intelligence targeting and conventional weapons capabilities of AI models
Anthropic’s Frontier Red Team developed new evaluations to measure AI capabilities in tactical intelligence targeting and conventional weapons development.
An alignment assessment of recent cybersecurity incidents
We present an alignment assessment of four incidents in which Claude models gained unauthorized access to real third-party systems.
Improving our alignment and security efforts
On July 30, we reported three incidents in which Claude models gained unauthorized access to real computer systems. We are conducting an in-depth analysis of both incidents, and planning to work with METR for an independent review. In the meantime, we’re sharing some of the changes we’ve made over the past month.
Investigating three real-world incidents in our cybersecurity evaluations
We found three cases where a Claude model reached the internet from an evaluation environment and accessed real systems. We share what happened and what we're changing.
Our position on open-weights models
Anthropic CEO Dario Amodei on open-weights models
Project Pilot: Can AI control a drone?
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
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.
A global workspace in language models
New interpretability research reveals an emergent mental workspace in Claude that holds internal thoughts that don’t appear in the model’s output.
Introducing Claude Sonnet 5
Sonnet 5 delivers frontier performance across coding, agents, and professional work at scale.
Measuring LLMs’ impact on N-day exploits
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 process of developing N-day exploits.