diff --git a/integration_tests/batch/test_batch.py b/integration_tests/batch/test_batch.py
index c42684026..d8bd7bfeb 100644
--- a/integration_tests/batch/test_batch.py
+++ b/integration_tests/batch/test_batch.py
@@ -64,7 +64,7 @@ def test_batch_openapi(page):
     expect(response).to_be_ok()
 
 
-def test_batch_request(unique_id, register_test_user, test_domain):
+def test_batch_request(unique_id, register_test_user, test_domain, docker_compose_exec):
     """A test via the Batch API should succeed."""
     request_data = {"type": "web", "domains": [test_domain], "name": unique_id}
 
@@ -129,6 +129,24 @@ def test_batch_request(unique_id, register_test_user, test_domain):
     response = requests.get(report_url, verify=False)
     assert response.status_code == 200, "test results should be publicly accessible without authentication"
 
+    # delete result files (this can happen in production when cleanup is done by a cron job)
+    docker_compose_exec("app", "sh -c 'rm -v /app/batch_results/*'")
+
+    # try to get batch results again after files have been deleted (this should trigger a new generation task)
+    results_response = requests.get(INTERNETNL_API + "requests/" + test_id + "/results", auth=auth, verify=False)
+    # expect error because result need to be generated again
+    assert results_response.status_code == 400
+    assert "Report is being generated." in results_response.text
+
+    # wait for report generation and batch to be done
+    wait_for_request_status(INTERNETNL_API + "requests/" + test_id, "done", timeout=60, auth=auth)
+
+    # get batch results again after starting generation
+    results_response = requests.get(INTERNETNL_API + "requests/" + test_id + "/results", auth=auth, verify=False)
+    print(f"{results_response.text=}")
+    results_response.raise_for_status()
+    print("api results JSON:", results_response.text)
+
 
 def test_batch_static_requires_no_auth():
     """Static files should be available without authentication for viewing batch results."""
diff --git a/interface/batch/util.py b/interface/batch/util.py
index c847f322b..48fd75977 100644
--- a/interface/batch/util.py
+++ b/interface/batch/util.py
@@ -273,6 +273,9 @@ def on_failure(exc, task_id, args, kwargs, einfo):
                 results = gather_batch_results_technical(user, batch_request, site_url)
                 save_batch_results_to_file(user, batch_request, results, technical=True)
 
+    request_lock_id = redis_id.batch_results_request_lock.id.format(batch_request.request_id)
+    cache.delete(request_lock_id)
+
 
 def gather_batch_results(user, batch_request, site_url):
     """
diff --git a/interface/batch/views.py b/interface/batch/views.py
index 9a2acda64..2774d4349 100644
--- a/interface/batch/views.py
+++ b/interface/batch/views.py
@@ -26,6 +26,7 @@
     register_request,
 )
 from internetnl import log
+from .util import request_already_generating
 
 
 @require_http_methods(["GET", "POST"])
@@ -70,9 +71,10 @@ def results(request, request_id, *args, technical=False, **kwargs):
         return bad_client_request_response("The request is not yet `done`.")
     else:
         if not batch_request.has_report_file():
+            if request_already_generating(batch_request.request_id):
+                return bad_client_request_response("Report is already being generated.")
             batch_async_generate_results.delay(user=user, batch_request=batch_request, site_url=get_site_url(request))
-            return bad_client_request_response("The request is not yet `done`.")
-
+            return bad_client_request_response("Report is being generated.")
         else:
             report_file = batch_request.get_report_file(technical)
             try:
diff --git a/interface/test/batch/test_batch.py b/interface/test/batch/test_batch.py
index 71df4bc2b..681ae1f05 100644
--- a/interface/test/batch/test_batch.py
+++ b/interface/test/batch/test_batch.py
@@ -3,6 +3,7 @@
 from django.core.cache import cache
 from checks.models import BatchRequest, BatchRequestType, BatchUser, BatchRequestStatus
 from interface.batch.util import get_request, register_request
+from interface.batch.views import results
 from django.test.client import RequestFactory
 import pytest
 from interface.batch.util import batch_async_generate_results
@@ -68,15 +69,27 @@ def test_batch_request_result_generation(db, client, mocker):
     batch_request.save()
 
     # if batch request is done, a batch_async_generate_results task should be put on the queue to generate the results
-    get_request(request, batch_request, test_user)
+    response = get_request(request, batch_request, test_user)
+    assert response.status_code == 200
+    assert json.loads(response.content)["request"]["status"] == "generating"
     assert batch_async_generate_results.delay.call_count == 1
 
     # there should not be an additional task put on the queue when one is already present
-    get_request(request, batch_request, test_user)
+    response = get_request(request, batch_request, test_user)
+    assert response.status_code == 200
+    assert json.loads(response.content)["request"]["status"] == "generating"
+    assert batch_async_generate_results.delay.call_count == 1
+
+    # when directly accessing results there should also not be a extra task added to the queue
+    assert (
+        results(request, batch_request.request_id, batch_user=test_user).status_code == 400
+    ), "should return `bad_client_request_response`"
     assert batch_async_generate_results.delay.call_count == 1
 
     # if the cache expires a new batch_async_generate_results task can be added
     lock_id = redis_id.batch_results_request_lock.id.format(batch_request.request_id)
     cache.delete(lock_id)
-    get_request(request, batch_request, test_user)
+    response = get_request(request, batch_request, test_user)
+    assert response.status_code == 200
+    assert json.loads(response.content)["request"]["status"] == "generating"
     assert batch_async_generate_results.delay.call_count == 2