Episode 335 - w/ Ron Perris - Secure Coding with AI okładka

Episode 335 - w/ Ron Perris - Secure Coding with AI

Episode 335 - w/ Ron Perris - Secure Coding with AI

Słuchaj bezpłatnie, bez subskrypcji

Zobacz szczegóły

Saisonale Sparangebote | 0,99 € pro Monat für die ersten 3 Monate

Danach 6,99 €/Monat. Bedingungen gelten. Monatlich kündbar.
In episode 335 of Absolute AppSec, Seth and Ken sit down with Ron Perris. Ron is Chief Technology Officer at Manicode Security, to discuss the integration of AI into secure software development and AppSec workflows. The conversation explores the rapid adoption of large language models (LLMs) and local agentic harnesses, examining how developers use them for code generation and how security teams manage the resulting system-level risks on developer workstations. Perris details his work with Manicode.ai, founded alongside Jim Manico, which addresses LLM code generation defects by embedding language- and framework-specific positive security guidance directly into the AI's context window. By leveraginglifecycle hooks within coding harnesses to inject tailored defensive rules (e.g., parameterized queries or framework-specific escaping), Manicode guides LLM token generation away from common vulnerabilities rather than relying solely on post-generation SAST checks. The hosts and Perris critique the industry's overreliance on dashboard-driven vulnerability counting, arguing that AppSec must move beyond managing repetitive bugs and focus on systematically eliminating flaw classes through secure framework defaults and context engineering. Finally, Perris reflects on his time at Reddit, the challenges of preventing recurring vulnerability classes at scale, and the potential for reviving practitioner-focused conferences like LocomocoSec to address modern AI-driven AppSec realities. Sponsored by Guardsquare (guardsquare.com).
adbl_web_anon_alc_button_suppression_t1
Brak recenzji