Lesson 891 of 1550
AI Ad-Targeting Audits: Catching Sensitive-Category Inferences
AI ad-targeting models can infer sensitive categories from innocuous signals — audit inference outputs, not just inputs.
Lesson map
What this lesson covers
Learning path
The main moves in order
- 1The premise
- 2proxy inference
- 3sensitive category
- 4GDPR Article 9
Concept cluster
Terms to connect while reading
Section 1
The premise
AI can probe ad-targeting models for proxy inferences of sensitive categories, but corrective action requires product and legal alignment.
What AI does well here
- Generate probe ad creatives that test for sensitive-category proxy targeting.
- Build an inference-audit log schema linking signals to categories.
What AI cannot do
- Decide which proxy inferences are commercially acceptable to allow.
- Remove a category from a deployed model without retraining oversight.
Key terms in this lesson
End-of-lesson quiz
Check what stuck
15 questions · Score saves to your progress.
Tutor
Curious about “AI Ad-Targeting Audits: Catching Sensitive-Category Inferences”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Adults & Professionals · 10 min
Bias Auditing in LLM Outputs: Seeing What the Model Can't
LLMs inherit the skews of their training data and RLHF feedback. Auditing for bias isn't a one-time test — it's an ongoing practice that belongs in every deployment.
Adults & Professionals · 40 min
Deepfake Detection: What Works, What Doesn't, and Why It Matters
AI-generated media has crossed the perceptual threshold where humans cannot reliably detect it. Detection tools help — but are in an arms race with generation.
Adults & Professionals · 11 min
Prompt Injection Defense: Protecting AI Systems From Malicious Inputs
Prompt injection is the SQL injection of the AI era — and it's already being exploited in production systems. Defending against it requires multiple layers, not a single fix.