Inkog：2026年AI智能体安全状况报告：基于500余项AI智能体开源项目的发现（英文版）（16页）.pdf

1、INKOGState of AI Agent Security 2026Findings from Scanning 500+Open-Source AI Agent Projects500+ReposScanned85.2%With Findings8050Total Findings63.4%CRITICAL/HIGHApril 2026 Inkog Security ResearchState of AI Agent Security 2026Findings from Scanning 500+Open-Source AI Agent ProjectsApril 2026|Inkog

2、Security ResearchExecutive SummaryWe scanned 500+open-source AI agent repositories using Inkogs static analysis engine the first automatedsecurity scanner purpose-built for AI agents.The results reveal a systemic security gap across the AI agentecosystem.Key findings:85.2%of repos contained at least

3、 one security finding63.4%had CRITICAL or HIGH severity vulnerabilities8050 total findings across 391 repositories(avg 20.59/repo)26.1%failed EU AI Act Article 14(Human Oversight)requirementsThe most common vulnerability:Infinite Loop(3312 instances)Most AI agents ship without basic security control

4、s that would be considered table stakes in traditional software.The gap between AI agent adoption and AI agent security is widening and with the EU AI Acts high-risksystem obligations taking effect in August 2026,the window for remediation is closing fast.Methodology:391 repos selected via 40 GitHub

5、 search queries targeting AI agent frameworks(LangChain,CrewAI,AutoGen,pydantic-ai,MCP servers,and 35+others).Minimum 20 stars,no forks.Scanned with Inkogv1.1.0 using comprehensive policy(all detectors enabled,no filtering).Full methodology in Section 3.1.The State of AI AgentsExplosive Growth,Minim

6、al SecurityAI agents have moved from research papers to production deployments at unprecedented speed.The global AIagent market reached approximately$5.25 billion in 2024 and is projected to scale past$10-15 billion by 2026,driven by high enterprise demand for workflow automation(MarketsandMarkets,G