Claude news February 2025

1 update from Claude and Anthropic in February 2025, from product launches to research papers, each with a short satori. note on what it means for a European business.

anthropic.com/research

3 Feb 2025

Constitutional Classifiers: Defending against universal jailbreaks

A defence system that cuts jailbreak success from 86 percent to 4.4 percent without breaking normal questions.

satori.

This is security work that actually works without making the model dull. Important if you are thinking of letting customers talk to your AI without them being able to trick it.

anthropic.com/research

Next month

We use analytics cookies to see which pages people read, and marketing cookies stay off unless you tick them under Details.

Cookie policy