diff --git a/subjects/ai/training/README.md b/subjects/ai/training/README.md
index db390147b..332311629 100644
--- a/subjects/ai/training/README.md
+++ b/subjects/ai/training/README.md
@@ -179,7 +179,7 @@ classifier.fit(X_train_scaled, y_train)
 
 ![alt text][logo_ex4]
 
-[logo_ex4]: ./w2_day4_ex4_q3.png 'ROC AUC '
+[logo_ex4]: ./w2_day4_ex4_q3.png "ROC AUC "
 
 - https://scikit-learn.org/stable/modules/generated/sklearn.metrics.plot_roc_curve.html
 
diff --git a/subjects/ai/training/audit/README.md b/subjects/ai/training/audit/README.md
index 75fe7a9a3..e654feac1 100644
--- a/subjects/ai/training/audit/README.md
+++ b/subjects/ai/training/audit/README.md
@@ -115,7 +115,7 @@ array([[37,  2],
 
 ![alt text][logo_ex4]
 
-[logo_ex4]: ../w2_day4_ex4_q3.png 'ROC AUC '
+[logo_ex4]: ../w2_day4_ex4_q3.png "ROC AUC "
 
 Having a 99% ROC AUC is not usual. The data set we used is easy to classify. On real data sets, always check if there's any leakage while having such a high ROC AUC score.