人工智能初创公司Anthropic展示了一种新技术，能够防止用户从其模型中获取有害内容。包括微软(Microsoft)和Meta在内的领先科技集团正在竞相寻找应对尖端技术带来危险的方法。 在周一发布的一篇论文中，总部位于旧金山的初创公司介绍了一种名为“宪法分类器 ...

2 月 6 日消息，据 The Information 报道，前 OpenAI 联合创始人 John Schulman 已离开 Anthropic。 Anthropic 首席科学官 Jared Kaplan 向路透社发布电子邮件并确认该 ...

2 月 5 日消息，在 Anthropic 公司的网站“客户案例”页面上，有大量报道称许多企业正在使用 Anthropic 的大语言模型 Claude，以帮助员工更有效地沟通。

The process of jailbreaking AI usually revolves around tricking the robot into not understanding ... no longer effective against Claude. Anthropic has employed a new system of Constitutional ...

Anthropic, the maker of the Claude AI chatbot, has an “AI policy” for applicants filling in its “why do you want to work here?” box and submitting cover letters (HT Simon Willison for the ...

Can you jailbreak Anthropic's latest AI safety measure? Researchers want you to try -- and are offering up to $20,000 if you succeed. Trained on synthetic data, these "classifiers" were able to ...

Lyft is partnering Anthropic to bring the startup's tech to its platform, starting with the customer care agent riders can access through the Lyft app.

Meanwhile, Anthropic already has extensive experience dealing with jailbreak attempts on Claude. The AI firm has devised a brand-new defense against universal AI jailbreaks called Constitutional ...