Researchers bypass 6 AI agent frameworks with a simple ‘Repeat-After-Me’ prompt
Researchers used a Repeat-After-Me prompt to make six AI agent frameworks run attacker-written tool calls.
66
Product
AutoGen
- Mentions · 7d0
- Mentions · 30d1
- All mentions2
- First seenAug 17, 2026
- Last mentioned18d ago
Timeline
2 articles · newest firstAI Agents Are Here. So Are the Threats.
Unit 42 demonstrates nine framework-agnostic attack scenarios against AI agents built with CrewAI and AutoGen, causing data leakage, credential theft and remote code execution.
58
Appears with
- CrewAIproduct · 2
- Unit 42vendor · 1
- Palo Alto Networksvendor · 1
- Agent Development Kitproduct · 1
- Université Montpellierorganization · 1
- LangChainvendor · 1
- CrewAIvendor · 1
- Strands Agentsproduct · 1
- Microsoftvendor · 1
- Model Context Protocolproduct · 1
- Llama 3.1 8Bai model · 1
- Metavendor · 1
- Googlevendor · 1
- Microsoft Agent Frameworkproduct · 1
- Prisma AIRSproduct · 1
- Pennsylvania State Universityorganization · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.