InfoQ Launches Specialized AI Security Cohort for Regulated Industries

InfoQ has opened applications for its AI Security & Privacy Engineering Program, a structured five-week online cohort designed for experienced engineers and architects navigating security and privacy challenges in AI deployments within heavily regulated sectors. Two intake periods are available—starting August 26 and October 14—with enrollment capped to practitioners holding at least five years of professional experience. Participants commit four hours weekly, with sessions led by Katharine Jarmul, author of Practical Data Privacy.

The Problem: AI Security in Isolation#

As artificial intelligence shifts from experimental sandbox projects into mission-critical business operations, teams face increasingly granular decisions: identifying which sensitive information flows into models, determining which threat scenarios merit dedicated defenses, placing appropriate guardrails and containment boundaries, and ensuring detection systems would surface problems faster than external parties would discover them. Yet most organizations tackle these questions largely alone, without industry peers or external expertise to validate their assumptions. This isolation leaves gaps in judgment and missed learning opportunities.

What the Program Offers#

The cohort pairs frameworks drawn from QCon conference talks with real-world application. Each week, participants examine an actual security or privacy challenge from their own production environments, then workshop their approaches with senior practitioners from different industries and companies—all within a confidential setting. Over five weeks, the curriculum spans sensitive data handling in AI pipelines, threat modeling methodologies (STRIDE, LINDDUN, Plot4AI), sandbox architecture and protective controls, monitoring and performance evaluation via platforms like Arize Phoenix, and compliance and audit processes. Teams culminate their learning in a documented risk assessment and mitigation strategy for an AI system; standout submissions are featured on InfoQ's platform.

Building Better Judgment#

Jarmul frames the discipline around organizational trust. As she notes, achieving AI systems that handle business-critical work or support core operations demands confronting privacy and security requirements head-on. A frequent organizational misstep involves teams focused on speed and automation overlooking fundamental information security practices, often betting that vendors will manage compliance obligations. Graduates of the program develop the capability to examine an AI architecture, spot its vulnerabilities, prioritize remediation efforts, and distinguish between threats that can be eliminated and those requiring ongoing observation.