Google 威脅情報團隊認定「可能由 AI 發現」的漏洞中,約 50% 可導致遠端程式碼執行;其他漏洞約為 26%。
零時差漏洞遭利用的增幅較整體漏洞利用趨勢溫和,防守方也須加快評估與修補已公開的已知漏洞。
How did AI affect the volume, types, risk levels, and exploitation of software vulnerabilities from January through August 2026, according tAI may help uncover vulnerabilities, but disclosure counts and observed exploitation measure different trends.
AI 提示詞
Create a landscape editorial hero image for this Studio Global article: How did AI affect the volume, types, risk levels, and exploitation of software vulnerabilities from January through August 2026, according t. Article summary: Google’s Threat Intelligence Group (GTIG) found that AI was accelerating vulnerability discovery and changing the mix of flaws being found, but the surge in disclosures did **not** mean a comparable share of flaws was be. Topic tags: general, general web, user generated. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fa
openai.com
Google 威脅情報團隊(GTIG)發現,2026 年 1 月至 8 月,軟體漏洞每月公開揭露數量增加逾一倍;同期觀察到遭攻擊者利用的漏洞數也上升,但幅度較小。GTIG 也指出,可能由 AI 協助發現的漏洞,類型與風險樣貌和其他漏洞有所不同,其中可導致遠端程式碼執行的比例較高。不過,這些資料無法釐清整體增幅有多少是 AI 單獨造成的。317