AI is changing how we code — and how attackers break systems.In this deep-dive conversation, Prabh and Mayank explore the new security challenges across frontend, backend, APIs, LLMs, supply chain, data pipelines, and infrastructure. From secure coding fundamentals to defending modern AI-driven systems, this episode is packed with real-world lessons developers and security teams can apply immediately.00:00 – 01:30 - Introduction, Guest welcome and his credentials 01:30 – 05:47 - Foundational Context and Core Premises05:47 – 08:30 - Importance of Secure AI coding08:30 – 09:50 - The Insecurity of LLM Outputs09:50 – 12:02 - Amplification of Attacks12:02 – 14:11 - Podcast Agenda14:11 – 21:25 - Front-end with example21:25 – 26:20 - Back-end26:20 – 31:10 – Database31:10 – 34:40 – Infrastructure34:40 – 37:04 – LLM and Prompt Injection37:04 – 39:33 - Output Filtering39:33 – 41:23 - Memory Attacks41:23 – 43:12 - Model Security43:12 – 54:24 - AI Supply Chain, Classics, and Future Threats54:24 – 57:40 - Incident prevention and always be ready57:40 – 01:03:40 – Typoglycemia01:03:40 – 01:09:40 - Due Diligence and Limited Visibility01:09:40 – 01:11:45 - Key challenges01:11:45 – 01:13:20 - Future Podcast Topic01:13:20 – 01:15:00 - End of the conversation by thanking Mayank Lau and looking forward to doing more Podcast.In this video, we break down everything developers and security teams must know about building secure systems in the age of AI. You’ll learn why AI-generated code isn’t automatically safe, how frontend vulnerabilities like XSS, SQL injection, sanitization failures, secrets exposure, and CORS issues still happen, and what it takes to design backend APIs that survive real attacks. We also go deep into LLM security — prompt injection, output manipulation, memory attacks, session poisoning, data theft, and how to secure agent pipelines the right way. You’ll understand the new AI supply chain risks, from unsafe data sources to third-party models, plus the importance of patching, hashing, and infrastructure monitoring for agentic systems. Finally, we discuss incident prevention and team mindset, covering MTTA/MTTR reduction, secure-team thinking, and why the right mindset protects you more than any tool.CISO talkshttps://www.youtube.com/playlist?list=PL0hT6hgexlYwPTD-wC3oFBe27VGEiizg1NIST Serieshttps://www.youtube.com/watch?v=VcC_KabV_Ho&list=PL0hT6hgexlYy0vBwMv0eteiyAxB48RQzy&pp=gAQBiAQBGRC Serieshttps://www.youtube.com/watch?v=mq_vSLHm4r0&list=PL0hT6hgexlYztA41j1bceTfVagP9mtq28&pp=gAQBISO 27001 Videohttps://www.youtube.com/watch?v=sQqJH2naU6I&t=1454s&pp=ygUeaXNvIDI3MDAxIGltcGxlbWVudGF0aW9uIHN0ZXBzISO 27001 Implementation Guidehttps://www.youtube.com/watch?v=GBfwk10Hh-o&pp=ygUeaXNvIDI3MDAxIGltcGxlbWVudGF0aW9uIHN0ZXBzGRC Practical Serieshttps://www.youtube.com/playlist?list=PL0hT6hgexlYztA41j1bceTfVagP9mtq28GRC Interviewhttps://www.youtube.com/playlist?list=PL0hT6hgexlYz1Usn1Nrnur6OzVoz59zylInternal Audithttps://www.youtube.com/playlist?list=PL0hT6hgexlYyNWBcGYfabwumCr0GKmLWv#SecureCoding #AISecurity #LLMSecurity #PromptInjection #BackendSecurity #FrontendSecurity #APISecurity #DevSecOps #AIAttacks #SupplyChainSecurity #MLSecurity #AgenticSystems #DeveloperSecurity #CyberSecurity2025 #vibecoding