# Some of the biggest financial institutions are now quietly experimenting with a powerful new AI system that can behave like a cyber attacker.

*fintech · news · 2026-04-13 · India Today*

## Key points

- Mythos, an AI model by Anthropic, is designed to mimic cyber attackers and reveal hidden vulnerabilities.
- Major US banks like JPMorgan, Goldman Sachs, Citi, and Morgan Stanley are quietly testing Mythos internally.
- Government officials have pressured banks to use advanced AI offensively for stress-testing, not just defense.
- Mythos demonstrated the ability to combine small flaws into major breaches, outperforming traditional testing methods.
- Access to Mythos is restricted under Project Glasswing, which includes Amazon and Apple alongside banks.

**Companies:** JPMorgan Chase, Goldman Sachs, Citigroup, Bank of America, Morgan Stanley, Amazon, Apple
**Countries:** United States

[Read the full story on India Today](https://www.indiatoday.in/technology/news/story/donald-trump-officials-push-banks-to-test-anthropic-mythos-ai-model-here-is-why-2895317-2026-04-13)

---

Canonical: https://newsio.io/n/ce38ee67-118d-4934-9fde-4920bb4ce023/some-of-the-biggest-financial-institutions-are-now-quietly-experimenting-with-a
Summarized by Newsio from India Today. https://newsio.io/how-it-works