Anthropic makes ‘jailbreak’ advance to stop AI models producing harmful results

Constitutional classifiers: New security system drastically reduces chatbot jailbreaks

A large team of computer engineers and security specialists at AI app maker Anthropic has developed a new security system aimed at preventing chatbot jailbreaks. Their paper is published on the arXiv preprint server.

Jailbreak Anthropic's new AI safety system for a $15,000 reward

In testing, the technique helped Claude block 95% of jailbreak attempts. But the process still needs more 'real-world' red-teaming.

Anthropic has a new security system it says can stop almost all AI jailbreaks

AI giant’s latest attempt at safeguarding against abusive prompts is mostly successful, but, by its own admission, still needs work.

Anthropic claims new AI security method blocks 95% of jailbreaks, invites red teamers to try

The new Claude safeguards have already technically been broken but Anthropic says this was due to a glitch — try again.

Anthropic makes ‘jailbreak’ advance to stop AI models producing harmful results

Leading tech groups including Microsoft and Meta also invest in similar safety systems

Anthropic has a new way to protect large language models against jailbreaks

This line of defense could be the strongest yet. But no shield is perfect.