Challenge #1 – PROTECT & SERVE the AI
Safeguard AI applications while ensuring seamless delivery and resilience
Background
AI adoption is growing rapidly, creating a need for strong defenses against attacks while maintaining high-performance and reliable delivery. Teams are challenged to build robust protections and delivery mechanisms for AI applications, ensuring security while safeguarding AI systems from malicious or unintended threats without compromising delivery speed or reliability
Key Focus Areas
​
-
AI Security: Defend against prompt injections, OWASP Top 10 vulnerabilities, and attacks targeting agents and LLM frameworks.
-
Agent Management: Safeguard interactions from both “good” and “malicious” agents.
-
Resilience & DDoS Protection: Mitigate DDoS attacks targeting AI services.
-
Reliable Application Delivery: Optimize delivery of AI applications using mechanisms like MCP load balancing, ensuring seamless performance and scalability.
-
MCP-Based Application Security: Implement security inspection within MCP protocols to protect WAF, API, and other application components against embedded attacks, ensuring robust security alongside reliable delivery.
Resources
-
Analysis Reports: About the evolving threats for AI Applications.
-
OWASP Top 10
-
Competitive Analysis: Access the latest technical competitive analysis conducted by our PMs.
-
Customer Feedback: Utilize insights from customer feedback collected by our Product Managers (PMs), Sales Engineers (SEs), and Professional Services (PS) teams.
Judging Criteria
45%
Business /Technological Viability
40%
Innovation, Creativity and Usability
15%
X - Factor
Outcomes
For all challenges the outcomes should be:
-
App - mockup or demo, algorithm, automation, tools, UX mockup, or any other means to demonstrate the suggested value of the selected challenge
-
Presentation: 3 minutes pitch: “Why?”- Need/Problem, “What?” – alternative solutions, suggested solution, the value, “How?” - implementation plan