AI in software development
MET-C04M01
Review generated code and catch hallucinated dependencies.
Code with an assistant without introducing flaws, phantom dependencies or exposed secrets.
Coming soon
Training by function
Software development
Coding assistants speed up your teams, and introduce flaws, hallucinated dependencies and secret leaks. This program makes critical review a reflex.
Why this training
Generated code is never ready without review.
An assistant can suggest a dependency that does not exist (typosquatting), reproduce a vulnerable pattern or prompt you to paste a real secret. Responsibility for the deliverable stays with the developer. This program anchors dependency verification, flaw hunting and secret hygiene.
The module dedicated to your function
Coming soon in the « AI by profession » specialization: this module is not available yet. The common core and the modules to go deeper below are available today.
The common core for everyone
The baseline reflexes your whole team shares, whatever the job.
Understanding generative AI
C01M01
Demystify the probabilistic engine and distinguish it from predictive AI.
Online10 min
Understand the fundamental principles and key mechanisms of how generative AI works.
RegisterHallucinations
C01M02
Identify fabricated information and how to verify it at the source.
Online13 min
Identify the causes of AI hallucinations and apply a rigorous method to guard against them.
RegisterWriting effective prompts
C01M03
Structure your prompts (Role + Context + Task + Format) and iterate.
Online8 min
Write structured prompts (RCTF) and iterate effectively to get precise results.
RegisterConfidential data and GenAI
C01M04
Sort your data and de-identify rigorously before every prompt.
Online8 min
Categorize work data and apply anonymization techniques to protect sensitive information.
RegisterApproved tools and shadow AI
C01M05
Recognize shadow AI and have tools approved by IT.
Online7 min
Identify the risks of shadow AI and favour officially approved tools.
RegisterYou remain responsible
C01M06
Apply "human-in-the-loop" and the "four eyes" method.
Online7 min
Take professional accountability for generated content and apply the principle of human verification.
RegisterGo deeper for your context
The modules most relevant to the risks specific to your function.
Prompt injection and AI agent risks
C03M03
Understand hidden instructions and apply least privilege.
Online9 min
Understand prompt-injection attacks and secure AI agents connected to your systems.
RegisterData leaks: what you must never paste
C03M01
Identify the blacklist never to submit to a public AI.
Online8 min
Describe how data entered into a consumer AI is handled by default and identify exfiltration risks.
RegisterIntellectual property and generated content
C04M03
Grasp the copyright grey zone and the risk of infringement.
Online7 min
Determine the legal status of AI-generated works and prevent infringement risks.
RegisterReady to equip your team?
Enroll your teams in our AI-literacy and governance training, from the common core to function-specific use cases.
Immediate access · 100% online · at your own pace