Baka Diop
CEO - Stealth AI
CEO
Paris, France
Actions
Baka Diop is a Global Ambassador for AI in France, award-nominated cybersecurity leader, and CEO of BluePinkSecurity. With 12 years of entrepreneurial experience and 9 years in cybersecurity, she drives innovative, AI-powered security solutions across Europe. A graduate of Paris-Saclay and ESIEA, she’s been named among the Top 50 Women in Cybersecurity. Baka is also a passionate advocate for diversity in tech and a sought-after international speaker.
Area of Expertise
Topics
How to empower Cybersecurity with AI?
In this keynote, attendees will gain a comprehensive understanding of the intersection between AI and application security. Balancing practical use cases, real-world examples, and strategic insight, this session will provide an insightful view into how AI can revolutionize application security.
Engineering and Securing AI-Powered Cyber Agents: Best Practices for Resilient Autonomous Defense
This session presents a practical approach to building and securing a modern cybersecurity stack with AI-driven agents.
As organizations move toward autonomous digital defense, each specialized agent ranging from vulnerability scanners to behavioral anomaly detectors is both a powerful solution and a potential attack surface.
In this talk, I’ll share real lessons from developing a modular agent-based architecture, highlighting how to:
- Design resilient AI agents as microservices, each responsible for a specific detection or response domain (e.g., vulnerabilities, misconfigurations, threats, containment).
- Enforce multi-layered, defense-in-depth strategies for agent security, including robust authentication, fine-grained access control, input validation, sandboxing, continuous monitoring, and incident response controls.
- Integrate external threat intelligence and vulnerability databases (MISP, OpenCTI, OTX, OSV, NVD, CVE Details, etc.) for adaptive and context-rich detection.
- Secure inter-agent communication and API interfaces, preventing data leakage and privilege escalation.
- Implement code-level best practices: secure development lifecycle, rigorous logging and auditing, role-based access, and automated vulnerability monitoring and patching for all agents.
Attendees will leave with actionable insights and a blueprint for securing AI agent architectures bridging cutting-edge automation with operational security rigor.
Key Takeaways :
- Strategies to minimize the attack surface of autonomous agents through layerez security and operational controls.
- Real examples of secure agent implementation: from input validation and API protection to logging, anomaly detection, and “kill switches” for fast incident response.
- Step-by-step integration of leading open-source intelligence and vulnerability platforms for maximal threat awareness.
- Design patterns to ensure every agent from deployment to remediation is secure by design and by operation.
This presentation targets engineering and security leaders building, deploying, or scaling AI-powered agents for critical cybersecurity tasks, emphasizing not just innovation but sustainable, enterprise-grade security at every layer.
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top