# Artificial intelligence giant Anthropic is rolling out an AI model it says is so powerful in its cybersecurity capabilities that it can only be used by specialists in the industry.

*business · news · 2026-04-07 · Australian Financial Review*

## Key points

- Anthropic has introduced a new AI model called Claude Mythos Preview for cybersecurity specialists.
- Claude Mythos Preview is considered too powerful for general use and restricted to industry experts.
- A website leak in March revealed the model early, impacting cybersecurity stock prices significantly.
- Anthropic has formed an alliance with major rivals and customers to combat malicious AI use.

**Companies:** CrowdStrike, Palo Alto Networks, Zscaler

[Read the full story on Australian Financial Review](https://www.afr.com/technology/anthropic-pits-bot-against-bot-in-ai-cyberwar-with-powerful-new-model-20260408-p5zm2f)

---

Canonical: https://newsio.io/n/66c94cfe-8514-4ff7-b760-dbe94f1fc35a/artificial-intelligence-giant-anthropic-is-rolling-out-an-ai-model-it-says-is-so
Summarized by Newsio from Australian Financial Review. https://newsio.io/how-it-works
