ainewsblitz.com

Breaking

Sakana AI Unveils Fugu-Cyber, Topping Security Benchmarks Held by OpenAI and Anthropic Rivals

Sakana AI Unveils Fugu-Cyber, Topping Security Benchmarks Held by OpenAI and Anthropic Rivals
  • Security
  • AI Agents
  • Foundation Models

Japan-based Sakana AI has released Fugu-Cyber, a cybersecurity-focused version of its Fugu orchestration system that posted leading scores on two real-world security benchmarks, edging past frontier offerings from larger labs. The model reached 86.9% on CyberGym and 72.1% on CTI-REALM, placing it at the top of both leaderboards and matching or surpassing dedicated cyber models such as OpenAI's GPT-5.5-Cyber and the Mythos Preview, according to the company's release announcement.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year