Hi folks,

The hardest part of security and privacy work for AI systems usually isn't the threat model. It's that most of us are doing it alone, within one company, with no one outside the team to check the decisions we make.

So I'm facilitating a five-week InfoQ online certification cohort built around exactly that. Each week, you take a security and privacy practice and apply it to your own work, then share with the group, in confidence, what worked and what didn't. Everyone in the room is a senior engineer or architect from a different company, and often a different industry, which is the whole point. You get to see how someone with the same problem, in a completely different context, is handling it.

To get real value from AI, you have to be able to trust it, and the moment you put it on business-critical work, you run straight into security and privacy requirements. Many senior engineers already feel this when working with coding assistants. What I keep seeing, though, is a strong bias toward action that skips the basics. So before you go faster, it's worth asking where your biggest risks are, which you can prevent, and, for the ones you can't, whether you'd find out first, before a third party does.

By the end of the cohort, you'll be able to look at an architecture and see where the risks are. You'll know how to prioritize them, which ones are preventable, and which ones you just need to watch. You'll be able to weigh the controls that address them: guardrails, agent sandboxing, and today's best practices in observability and evaluation. You leave with a working foundation in security and privacy engineering for AI systems, and a good number of places to go deeper on your own.

Five weeks, a few hours per week, live sessions starting August 26 and October 14 (PM CEST). Enrollment is limited to senior engineers and software architects with a minimum of five years' experience. 

Most companies reimburse for professional development, and there's a "convince your boss" template if you need it.

See the full syllabus and reserve your place.

Katharine Jarmul
Author of Practical Data Privacy (O'Reilly), and facilitator, InfoQ Certified AI Security & Privacy Engineering Program

P.S. Two other InfoQ Online Certification Program cohorts are running this summer:

• AI Engineering with Hien Luu, author of MLOps with Ray (starts July 25). Getting AI systems past the prototype stage: AI-native engineering, RAG and context pipelines, agents, platform and infrastructure, evals, and reliability.

• Architecture with Luca Mezzalira, author of Building Micro-Frontends (starts August 13, with September and October groups). The sociotechnical side of architecture: trade-offs, decentralized decisions, platform engineering, and AI architecture calls.

P.P.S. If you're thinking about attending QCon San Francisco (Nov 16-18, 2026) or QCon London (April 13-16, 2027), early bird pricing ends July 14.

------------------------------------------------------------

No longer want to receive emails about the Online InfoQ Certified Program? Unsubscribe.

C4Media Inc. (QCon & InfoQ),

705-2267 Lake Shore Blvd. West

Toronto, Ontario, Canada

M8V 3X2