CyberArena symmetric attack-defense evaluation of 12 models

1. 06:20
   WeChat · Synced (Jiqizhixin)Selected

   Cybersecurity's AlphaGo moment, or starting from CyberArena's 466 agent matches

   The research team built CyberArena, a symmetric attack-defense evaluation, having 12 frontier model and harness combinations simultaneously attack opponents and maintain their own services across 22 batches and 466 valid matches. GPT-5.6 Sol ranked first with 78.6 points, Claude Opus 5 second with 73.8, and GLM-5.3 third with 68.5. Each match had 3 rounds of 20 minutes each; patches deployed by the defending side persist, and shutting down services or harming normal business incurs point deductions. The paper says that for the 30 problems all three models could solve alone, after entering two-way adversarial play the round-3 attack capture rates fell to 52%, 37% and 35% respectively, indicating a gap between static CTF scores and dynamic attack-defense performance. An audit of the 466 matches found 2,397 errors, and in 23% of matches the agent claimed the fix was complete based only on logs from before the patch was deployed.

1. 03:54
   X@evfuturenews· Discussion

   CyberArena-მ 12 AI აგენტი შეაფასა 40 გამოწვევაზე ჩატარებულ 466 ვალიდურ, ცოცხალ თავდაცვით-შემტევ მატჩში. ანგარიშში GPT-5…(opens the original post in a new tab)
2. 03:53
   X@evfuturenews· Discussion

   CyberArena tested 12 AI agents across 466 valid live attack-and-defense matches on 40 challenges. GPT-5.6 Sol led the r…(opens the original post in a new tab)

1. 18:31
   WebSynced (Jiqizhixin)· Media coverage

   Cybersecurity's AlphaGo moment, or starting from these 466 agent matches(opens the original post in a new tab)
2. 18:00
   WeChat · Synced (Jiqizhixin)Selected

   Cybersecurity's AlphaGo moment, or starting from CyberArena's 466 agent matches

   A research team from 2077AI, UC Berkeley RDI and other institutions launched the symmetric attack-defense evaluation platform CyberArena; 12 frontier model and harness combinations went through 22 batches and 466 valid matches. OpenAI's GPT-5.6 Sol ranked first with 78.6 points, Anthropic's Claude Opus 5 second with 73.8 points, and Zhipu's open-source model GLM-5.3 third with 68.5 points. The platform uses an attack-defense capture-the-flag format, with 3 rounds per match; the agent must simultaneously break into the opponent's services and defend the availability of its own business. Comparative experiments showed that on 30 problems that GPT-5.6 Sol, Claude Opus 4.8 and GLM-5.3 could all fully solve in single-player mode, after entering two-way confrontation the attack capture rates in round 3 fell to 52%, 37% and 35% respectively. The paper is arXiv 2610.19005, and the code has been open-sourced on GitHub.

3. 11:58
   X@AIbreakingnew· Discussion

   AI agents now face off live: each attacks the rival service while patching its own. Over 466 matches, GPT-5.6 Sol in Co…(opens the original post in a new tab)

Current attention 42·Peak in the comparable period 43 (Nov 6 03:00)·Change over the comparable past 24 hours 0%