Virtual Briefing: How to Secure AI Agents Before They Go Rogue

As AI agents become more prevalent, the risk of indirect prompt injection is rising as a leading threat in the agentic AI landscape. By 2029, over 50% of successful cybersecurity attacks against AI agents will exploit access control issues, with prompt injection as a primary attack vector. Cybersecurity teams must secure all AI agents their business units are buying or building to effectively mitigate risk. Join this invitation-only, complimentary virtual briefing as Gartner experts discuss the top security risks associated with LLM-based AI agents. Gain practical strategies for managing rogue agent threats and discover best practices for securing MCP implementations. By understanding and addressing these risks, cybersecurity teams will be better equipped to protect their organizations against evolving threats in the AI landscape. You will walk away from this session with answers to your vital questions and recommended actions to help you achieve your goals.

• Examine AI agent discovery and security posture management

• Learn effective agent defense strategies

• Explore critical priorities for security and agent skills protection

Return to this web page to watch the virtual briefing. Contact us at gartnervirtualbriefings@gartner.com with questions about viewing this virtual briefing.