← Back to news

Multi-turn attacks broke AI models 88% of the time

When Cisco ran 6,986 multi-turn attacks against 15 flagship models, attackers who adapted across the conversation broke through as often as 88.3% of the time. Amy Chang, Cisco's head of AI threat intelligence and security research, brought that finding to the agentic security panel at VB Transf

• Multi-turn attacks broke AI models 88% of the time — single-turn testing missed it, Cisco AI security lead warns at VB Transform 2026

  VentureBeat 23 Jul 2026

  When Cisco ran 6,986 multi-turn attacks against 15 flagship models, attackers who adapted across the conversation broke through as often as 88.3% of the time. Amy Chang, Cisco's head of AI threat intelligence and security research, brought that finding to the agentic security panel at VB Transf

• ServiceNow Goes All-In on Cybersecurity As AI Risk Keeps CEOs Awake at Night
  1 source 23 Jul 2026
• CISA orders urgent action on actively exploited Langflow RCE flaw
  1 source 22 Jul 2026
• The agent security gap: 54% of enterprises have already had an AI agent incident, and most still let agents share credentials
  1 source 23 Jul 2026
• Rogue OpenAI Agent Hacks Hugging Face: Is Your CX Team Prepared for Runaway Autonomous Agents?
  1 source 22 Jul 2026
• JadePuffer agentic attacks now target AI model data with ransomware
  1 source 20 Jul 2026