InfoQ Opens AI Security & Privacy Engineering Cohort for Regulated Industries

InfoQ has opened enrollment for the AI Security & Privacy Engineering Program, a five-week online cohort for senior engineers and architects responsible for security and privacy for AI systems in regulated industries. Two cohorts are scheduled, one starting August 26 and one starting October 14, each limited to practitioners with at least five years of experience. Sessions run four hours a week and are facilitated by Katharine Jarmul, author of Practical Data Privacy (O'Reilly).

Security and privacy work for production AI is largely confined within a single company, with few external voices to check the decisions. As AI moves from experiments into business-critical use, the questions get more specific. Teams have to work out what sensitive data is reaching the model, which threats are worth modeling, where guardrails and sandboxes belong, and whether their observability would catch a failure before a third party does. Most of that happens with little to compare against, and the cohort presents scenarios, trade-offs, and decisions to a confidential peer group. Each week, participants take a framework from a QCon talk and apply it to a real security or privacy decision from their own work, then discuss what worked, what failed, and what is still uncertain with senior engineers and architects from other companies and industries.

Katharine frames the stakes around trust. "Want AI to help with business-critical tasks or support the most meaningful work at your organization? Then you're going to run directly into privacy and security requirements," she said. She points to a common blind spot. Teams biased toward action and automation skip the basics of information security and privacy, often assuming a vendor will handle compliance. After the cohort, participants should be able to look at an AI architecture, identify the risks, decide where to start, and determine which risks can be prevented and which only require monitoring.

The five-week cohort covers working with sensitive data in AI workflows, through threat modeling and hands-on red teaming (using methods such as STRIDE, LINDDUN, and Plot4AI), into controls and sandboxes, then observability and evaluation with tools like Arize Phoenix, and finally governance and auditing. Each working group produces a documented risk assessment and mitigation report for an AI product architecture as its capstone, and the best are published on InfoQ.

The program costs USD 1,470 per cohort, and most companies reimburse professional development. A "convince your boss" template is available for anyone who needs to request approval.

Two other InfoQ Online Certification Programs are running alongside it. AI Engineering with Hien Luu, author of MLOps with Ray, starts July 25 and covers getting AI systems past the prototype stage, including RAG and context pipelines, agents, evaluation, and reliability. Architecture with Luca Mezzalira, author of Building Micro-Frontends, starts August 13 and covers the sociotechnical side of architecture, including trade-offs and communication, decentralized decision-making, platform engineering, and AI architecture decisions.

The full syllabus and registration are on the AI Security & Privacy Engineering Program page.