News
Introducing Claude Corps
13+ hour, 39+ min ago (358+ words) We're launching Claude Corps, a national fellowship program for people early in their careers who are passionate about extending the benefits of AI to communities across America. We'll teach 1, 000 fellows how to use Claude well, match them with nonprofits across…...
What we learned mapping a year's worth of AI-enabled cyber threats
1+ week, 1+ day ago (531+ words) As AI transforms the nature of and methods behind cyberattacks, how well do the techniques and frameworks used by the security community hold up? There were three main conclusions from our analysis: Below we provide a summary of each of…...
Coding agents in the social sciences
2+ week, 1+ day ago (693+ words) The human sciences are shifting: for the first time, core research tasks can be handed off to machines. AI chatbots increasingly contribute to scientific research, including in the most prestigious publications and in the social sciences. This has spurred optimism…...
Natural Language Autoencoders
1+ mon, 4+ day ago (836+ words) When you talk to an AI model like Claude, you talk to it in words. Internally, Claude processes those words as long lists of numbers, before again producing words as its output. These numbers in the middle are called activations'and…...
Emotion concepts and their function in a large language model
2+ mon, 1+ week ago (30+ words) Anthropic Application error: a client-side exception has occurred while loading www. anthropic. com (see the browser console for more information)....
The persona selection model
3+ mon, 2+ week ago (558+ words) Why would AI assistants behave like they're human? A natural guess might be that AI developers train them to do so. There's some truth to this: Anthropic trains Claude to chat conversationally with users, to respond warmly and empathetically, and…...
Disempowerment patterns in real-world AI usage
4+ mon, 2+ week ago (1238+ words) AI assistants are now embedded in our daily lives'used most often for instrumental tasks like writing code, but increasingly in personal domains: navigating relationships, processing emotions, or advising on major life decisions. In the vast majority of cases, the influence…...
Detecting and countering misuse of AI: August 2025
9+ mon, 1+ week ago (528+ words) We've developed sophisticated safety and security measures to prevent the misuse of our AI models. But cybercriminals and other malicious actors are actively attempting to find ways around them. Today, we're releasing a report that details how. Our Threat Intelligence…...