Training Claude on copyrighted books it purchased was fair use, but piracy wasn't, the judge ruled.

Court documents reveal Anthropic destroyed millions of physical books to train Claude AI, creating new legal precedent for ...

New research shows Claude chats often lift users’ moods. Anthropic explores how emotionally supportive AI affects behavior, ...

A new Anthropic report shows exactly how in an experiment, AI arrives at an undesirable action: blackmailing a fictional ...

New research from Anthropic suggests that most leading AI models exhibit a tendency to blackmail, when it's the last resort in certain tests.

In a test case for the artificial intelligence industry, a federal judge has ruled that AI company Anthropic didn’t break the ...

New research shows that as agentic AI becomes more autonomous, it can also become an insider threat, consistently choosing ...

Well, Anthropic received a legal win this week when a court ruled that it didn’t break the law by training Claude on the ...

Anthropic’s Claude Opus 4 turned to blackmail 96% of the time, while Google’s Gemini 2.5 Pro had a 95% blackmail rate. OpenAI’s GPT-4.1 blackmailed the executive 80% of the time, and ...

Artificial intelligence company Anthropic has released new research claiming that artificial intelligence (AI) models might ...

Google has released its terminal-based Gemini CLI agent which is open-source and offers an expanded usage limit for the ...

Anthropic research reveals AI models from OpenAI, Google, Meta and others chose blackmail, corporate espionage and lethal actions when facing shutdown or conflicting goals.