diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index b01a02bf5e78a829a82446b1f419f20f18f5d261..e2230ceafce0bb041cf0075923eef2d62e1821fe 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -449,17 +449,21 @@ codec-msan:
     - *print-common-info
     - make clean
     - make -j CLANG=1
-    - python3 scripts/self_test.py -z console --create | tee test_output.txt
-    - run_errors=$(cat test_output.txt | grep -ic "run errors") || true
-    - if [ $run_errors != 0 ] ; then echo "Run errors in self_test.py with Clang memory-sanitizer"; exit 1; fi
+    - testcase_timeout=180
+    - exit_code=0
+    - python3 -m pytest tests/codec_be_on_mr_nonselection/test_param_file.py -v --update_ref 1 -m create_ref --html=report.html --self-contained-html --junit-xml=report-junit.xml --testcase_timeout=$testcase_timeout --ref_encoder_path ./IVAS_cod --ref_decoder_path ./IVAS_dec || exit_code=$?
+    - if [ $exit_code != 0 ] ; then echo "Run errors found by Clang memory-sanitizer"; exit 1; fi
   artifacts:
     name: "mr-$CI_MERGE_REQUEST_IID--sha-$CI_COMMIT_SHORT_SHA--stage-$CI_JOB_STAGE--results"
     expire_in: 1 week
     when: always
     paths:
-      - scripts/ref/logs/
-      - test_output.txt
+      - report-junit.xml
+      - report.html
     expose_as: "msan selftest results"
+    reports:
+      junit:
+        - report-junit.xml
 
 # code selftest testvectors with address-sanitizer binaries
 codec-asan:
@@ -472,17 +476,21 @@ codec-asan:
     - *print-common-info
     - make clean
     - make -j CLANG=2
-    - python3 scripts/self_test.py -z console --create | tee test_output.txt
-    - run_errors=$(cat test_output.txt | grep -ic "run errors") || true
-    - if [ $run_errors != 0 ] ; then echo "Run errors in self_test.py with Clang address-sanitizer"; exit 1; fi
+    - testcase_timeout=180
+    - exit_code=0
+    - python3 -m pytest tests/codec_be_on_mr_nonselection/test_param_file.py -v --update_ref 1 -m create_ref --html=report.html --self-contained-html --junit-xml=report-junit.xml --testcase_timeout=$testcase_timeout --ref_encoder_path ./IVAS_cod --ref_decoder_path ./IVAS_dec || exit_code=$?
+    - if [ $exit_code != 0 ] ; then echo "Run errors found by Clang address-sanitizer"; exit 1; fi
   artifacts:
     name: "mr-$CI_MERGE_REQUEST_IID--sha-$CI_COMMIT_SHORT_SHA--stage-$CI_JOB_STAGE--results"
     expire_in: 1 week
     when: always
     paths:
-      - scripts/ref/logs/
-      - test_output.txt
+      - report-junit.xml
+      - report.html
     expose_as: "asan selftest results"
+    reports:
+      junit:
+        - report-junit.xml
 
 # code selftest testvectors with address-sanitizer binaries
 codec-usan:
@@ -495,17 +503,21 @@ codec-usan:
     - *print-common-info
     - make clean
     - make -j CLANG=3
-    - UBSAN_OPTIONS=suppressions=scripts/ubsan.supp,report_error_type=1 python3 scripts/self_test.py -z console --create
-    - grep_exit_code=0
-    - grep UndefinedBehaviorSanitizer scripts/ref/logs/* || grep_exit_code=$?
-    - if [ $grep_exit_code != 1 ] ; then echo "Run errors in self_test.py with Clang undefined-behavior-sanitizer"; exit 1; fi
+    - testcase_timeout=180
+    - exit_code=0
+    - UBSAN_OPTIONS=suppressions=scripts/ubsan.supp,report_error_type=1 python3 -m pytest tests/codec_be_on_mr_nonselection/test_param_file.py -v --update_ref 1 -m create_ref --html=report.html --self-contained-html --junit-xml=report-junit.xml --testcase_timeout=$testcase_timeout --ref_encoder_path ./IVAS_cod --ref_decoder_path ./IVAS_dec || exit_code=$?
+    - if [ $exit_code != 0 ] ; then echo "Run errors found by Clang undefined-behavior-sanitizer"; exit 1; fi
   artifacts:
     name: "mr-$CI_MERGE_REQUEST_IID--sha-$CI_COMMIT_SHORT_SHA--stage-$CI_JOB_STAGE--results"
     expire_in: 1 week
     when: always
     paths:
-      - scripts/ref/logs/
+      - report-junit.xml
+      - report.html
     expose_as: "usan selftest results"
+    reports:
+      junit:
+        - report-junit.xml
 
 # test renderer executable
 renderer-smoke-test:
diff --git a/tests/conftest.py b/tests/conftest.py
index cccae2e90662e37b3d34ec772c588074ce5b8d58..4250c2fe876e941bd07e10e3ba9363728c237c75 100644
--- a/tests/conftest.py
+++ b/tests/conftest.py
@@ -292,6 +292,8 @@ class EncoderFrontend:
             pytest.fail(
                 f"{self._type} encoder terminated with a non-0 return code: {self.returncode}"
             )
+        if self.stderr and "UndefinedBehaviorSanitizer" in self.stderr:
+            pytest.fail("Undefined Behaviour runtime error encountered")
 
     def _check_run(self):
         if self.returncode is not None:
@@ -485,6 +487,8 @@ class DecoderFrontend:
             pytest.fail(
                 f"{self._type} decoder terminated with a non-0 return code: {self.returncode}"
             )
+        if self.stderr and "UndefinedBehaviorSanitizer" in self.stderr:
+            pytest.fail("Undefined Behaviour runtime error encountered")
 
     def _check_run(self):
         if self.returncode is not None: