Back to index
Other10+ hours of video plus local labs; 2-day practical exam and 2 days for the report·$249 for courses + exam with one free retake (AI 100 is free; Academy from $29.99/month)

TCM Security AI Hacking 101 and Practical AI Pentest Associate (PAPA)

4.3

The best value hands-on AI pentesting certification I have looked at. For $249 you get the training, local labs and an exam that makes you attack an agentic AI app and write a report, with a free retake. It is associate level and does not have the recognition of OffSec or SANS, but the format is right.

What We Liked

  • $249 covers the course, labs, an exam attempt and a free retake, all valid for 12 months
  • Fully practical exam against an agentic AI application with zero flags and zero multiple-choice questions
  • Report writing is graded, which is what a real AI pentest engagement is actually judged on
  • No proctoring software, and all tools including AI tools are allowed, which matches how people really work
  • AI 100: Fundamentals is free, so you can check the teaching style before paying
  • The certification does not expire at the moment, so there are no renewal fees

What Could Be Better

  • Associate level by design; experienced web pentesters may find the course material light
  • Brand-new credential with little recognition among HR filters compared with OSCP-style names
  • Labs run locally, so you need a machine that can handle virtual machines and models
  • The field moves fast, and a single 6-hour course will date quickly unless TCM keeps updating it

Detailed review

TCM Security made its name with cheap, practical certifications like the PNPT, and it has applied the same formula to AI. The track has three parts. AI 100: Fundamentals is a free four-hour course on how large language models and neural networks work. AI Hacking 101 is about six hours on attacking LLM-based applications, covering prompt injection and jailbreaking, sensitive data exfiltration through RAG, improper output handling and the rest of the usual OWASP LLM list.

Both are taught by Andrew Bellini, an electrical engineer turned ethical hacker who is behind all of TCM's AI material. A third course, Securing AI Applications, covers the defensive side. The PAPA certification is what makes this worth reviewing. For $249 you get both courses for 12 months, local labs, one exam attempt and one free retake. The exam gives you two days to pentest an agentic AI application in a dedicated lab, then two more days to write a report.

There are no flags and no multiple-choice questions. TCM's line is that it does not want to profit from failure, which is why the retake is included. I like this format a lot. Most AI security credentials I have reviewed are multiple-choice tests about vocabulary. Testing an agent end to end and then writing up findings for a client is the actual job.

Allowing AI tools in the exam is also the honest choice, since everyone doing this work uses them. The weaknesses are about level and recognition. This is an associate exam, and if you have years of web testing behind you, much of the course will feel familiar with an LLM twist. The name PAPA also means nothing to most recruiters yet. Compared with OffSec's AI-300, which starts around $1,749, or SANS courses that run to thousands, PAPA is a fraction of the price, so the trade-off is reasonable. One more thing to plan for: the labs are local, so you need a laptop that can run VMs comfortably.

Check that before you buy, because your 12-month clock starts at purchase. Overall, this is the AI pentesting cert I would point most people to first. It is cheap, practical, honest about its level and graded on a report rather than trivia.

[ final ]

The verdict.

If you already do web or network pentesting and want to add LLM and agent testing, buy this before anything more expensive. Start with the free AI 100 course, then take PAPA. Save OffSec's AI-300 or SANS for when an employer is paying.