diff --git a/CHANGELOG.md b/CHANGELOG.md
index 69b1ba743..1cf797076 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,6 +12,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - It is now possible to tell whether `UserStartedSpeakingFrame` or
   `UserStoppedSpeakingFrame` have been generated because of emulation frames.
 
+### Fixed
+
+- Fixed an issue that would cause `SegmentedSTTService` based services
+  (e.g. `OpenAISTTService`) to try to transcribe non-spoken audio, causing
+  invalid transcriptions.
+
 ## [0.0.61] - 2025-03-26
 
 ### Added
diff --git a/src/pipecat/services/ai_services.py b/src/pipecat/services/ai_services.py
index a78c268dd..97aad0d40 100644
--- a/src/pipecat/services/ai_services.py
+++ b/src/pipecat/services/ai_services.py
@@ -1048,9 +1048,14 @@ class SegmentedSTTService(STTService):
             await self._handle_user_stopped_speaking(frame)
 
     async def _handle_user_started_speaking(self, frame: UserStartedSpeakingFrame):
+        if frame.emulated:
+            return
         self._user_speaking = True
 
     async def _handle_user_stopped_speaking(self, frame: UserStoppedSpeakingFrame):
+        if frame.emulated:
+            return
+
         self._user_speaking = False
 
         content = io.BytesIO()
@@ -1068,7 +1073,7 @@ class SegmentedSTTService(STTService):
         self._audio_buffer.clear()
 
     async def process_audio_frame(self, frame: AudioRawFrame, direction: FrameDirection):
-        # If the user is speaking the audio buffer will keep growin.
+        # If the user is speaking the audio buffer will keep growing.
         self._audio_buffer += frame.audio
 
         # If the user is not speaking we keep just a little bit of audio.