Anthropic’s New Artificial Intelligence Models Tackle Cybersecurity Challenges

Introducing Claude Fable 5 and Claude Mythos 5

Anthropic has launched two new artificial intelligence models: Claude Fable 5 and Claude Mythos 5. These models exceed the capabilities of the previous Mythos Preview.

Cybersecurity Focus

With increasing concerns over cybersecurity, the models incorporate enhanced security features to prevent potential misuse by bad actors. Claude Mythos 5 is currently available only to select industry partners under Project Glasswing, aimed at enhancing collaborative security measures.

Guardrails Implementation

The newly devised content moderation techniques will block critical queries, ensuring sensitive questions related to hacking and vulnerabilities are routed to an older AI model, Claude Opus 4.8. This strategy is designed to mitigate risks associated with AI vulnerabilities and potential hacking tools.

Collaboration with the Government

• Anthropic collaborates with the US government for rollout.
• Focused on improving security while maintaining user accessibility.

Diane Penn, head of product management, noted that they have been addressing the handling of vulnerabilities since before the April release of Mythos. They assess user feedback continuously to refine their approach.

Pricing and Access

Pricing for Claude Fable 5 and Claude Mythos 5 is set at $10 per million input tokens and $50 per million output tokens. This pricing reflects the advanced capabilities provided compared to earlier models.

Future Prospects

As the AI landscape evolves, Anthropic aims to expand access while ensuring adequate safeguards are in place. This measured approach is crucial in addressing inherent vulnerabilities associated with high-performance AI models.