The UK Government’s Mythos AI Tests Help Separate Cybersecurity Threat from Hype
- The UK government’s AI Security Institute (AISI) has conducted an initial evaluation of Anthropic’s Mythos Preview model, adding independent public verification to Anthropic’s claims.
- While Mythos isn’t significantly different in its cybersecurity capabilities compared to other recent models, AISI highlights its potential for chaining tasks into complex multistep attacks, which could distinguish it from previous models.
- AISI’s evaluations show that Mythos Preview can complete over 85 percent of the low-level “Apprentice” tasks in Capture the Flag challenges, demonstrating significant progress and setting a new benchmark in cybersecurity AI testing.
Originally published at arstechnica.com. Curated by AI Maestro.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.