Unverified70% confidenceFactExact time
Models such as GPT-4 demonstrated vulnerability discovery and exploitation capabilities in CTF competitions during 2022–2023.
2
Sources
70%
Confidence
Long-term
Relevance
8/6/2026
First Seen
Sources
AI网络安全评估事件复盘：第三方测试中的安全治理启示
Related Claims
UnverifiedThe release of ChatGPT at the end of 2022 and GPT-4's demonstration of multi-step reasoning capabilities in 2023 invalidated the capability prediction curves that had previously been widely accepted in the industry79% similarVerifiedMultiple academic studies in 2024 showed that models such as GPT-4 are able to identify certain types of software vulnerabilities in controlled environments, with efficiency approaching that of a junior security researcher.78% similarUnverifiedResearch in 2024 showed that models such as GPT-4 and Claude, assisted by auxiliary tools, are already capable of independently solving approximately 30%–40% of medium-difficulty CTF challenges.78% similarUnverifiedGPT-4 is capable of solving medium-difficulty binary exploitation challenges in CTF competitions, whereas GPT-3 is almost unable to complete full exploit development.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/699133API
curl https://kongchang.com/api/v1/knowledge/claims/699133MCP
get_claim(id=699133)