diff --git a/cli/medperf/commands/benchmark/update_associations_poilcy.py b/cli/medperf/commands/benchmark/update_associations_poilcy.py
index 37108334d..8ac1e8211 100644
--- a/cli/medperf/commands/benchmark/update_associations_poilcy.py
+++ b/cli/medperf/commands/benchmark/update_associations_poilcy.py
@@ -112,6 +112,21 @@ def validate_emails(self):
         if self.model_emails is not None:
             self.model_emails = validate_and_normalize_emails(self.model_emails)
 
+    def warn_if_empty_allow_list(self):
+        # self.dataset_emails is None means the field wasn't part of this
+        # submission at all (existing allow list, if any, stays untouched);
+        # only an explicitly-submitted empty list ([]) warrants a warning.
+        if self.dataset_mode == "ALLOWLIST" and self.dataset_emails == []:
+            config.ui.print_warning(
+                "Dataset allow list is empty: no dataset associations will be "
+                "auto-approved until you add emails to the allow list."
+            )
+        if self.model_mode == "ALLOWLIST" and self.model_emails == []:
+            config.ui.print_warning(
+                "Model allow list is empty: no model associations will be "
+                "auto-approved until you add emails to the allow list."
+            )
+
     def update(self):
         if all(
             [
@@ -122,6 +137,7 @@ def update(self):
             ]
         ):
             return
+        self.warn_if_empty_allow_list()
         body = {}
         if self.dataset_emails is not None:
             body["dataset_auto_approval_allow_list"] = self.dataset_emails
diff --git a/cli/medperf/commands/mlcube/grant_access.py b/cli/medperf/commands/mlcube/grant_access.py
index 131b60db7..46696ab61 100644
--- a/cli/medperf/commands/mlcube/grant_access.py
+++ b/cli/medperf/commands/mlcube/grant_access.py
@@ -61,6 +61,11 @@ def get_approval(self):
             f"registered in Benchmark (UID: {self.benchmark_id}) access to "
             "your Model.\n"
         )
+        if not self.allowed_emails:
+            msg += (
+                "No allow list was provided, so ALL eligible Data Owners will "
+                "be granted access (no email filtering).\n"
+            )
 
         if not self.approved and not approval_prompt(msg):
             raise CleanExit("Access granting operation cancelled")
@@ -104,7 +109,7 @@ def prepare_certificates_list(self):
         self.cert_user_info = cert_user_info
 
     def filter_certificates(self):
-        if self.allowed_emails is None:
+        if not self.allowed_emails:
             return
         logging.debug("Filtering certificates based on allowed emails list")
         filtered_certificates = []
diff --git a/cli/medperf/tests/commands/mlcube/test_grant_access.py b/cli/medperf/tests/commands/mlcube/test_grant_access.py
index bb6367d72..c62aaa876 100644
--- a/cli/medperf/tests/commands/mlcube/test_grant_access.py
+++ b/cli/medperf/tests/commands/mlcube/test_grant_access.py
@@ -217,9 +217,11 @@ def test_filter_certificates_when_empty(grantaccess):
     }
     grantaccess.allowed_emails = []
 
-    # Act & Assert
-    with pytest.raises(CleanExit):
-        grantaccess.filter_certificates()
+    # Act
+    grantaccess.filter_certificates()
+
+    # Assert
+    assert [cert.id for cert in grantaccess.certificates] == [1, 2]
 
 
 def test_verify_certificates_filters_invalid_certs(mocker, grantaccess):
diff --git a/cli/medperf/ui/web_ui.py b/cli/medperf/ui/web_ui.py
index d5ee40578..dd7f8038a 100644
--- a/cli/medperf/ui/web_ui.py
+++ b/cli/medperf/ui/web_ui.py
@@ -1,3 +1,4 @@
+import threading
 from queue import Queue
 from contextlib import contextmanager
 from yaspin import yaspin
@@ -16,6 +17,39 @@ def __init__(self):
         self.task_id = None
         self.events_manager = EventsManager()
         self.global_events_manager = GlobalEventsManager()
+        self._silenced = threading.local()
+
+    @property
+    def _is_silenced(self) -> bool:
+        return getattr(self._silenced, "value", False)
+
+    @property
+    def _captured_messages(self):
+        return getattr(self._silenced, "captured", None)
+
+    @contextmanager
+    def capture(self):
+        """Run a block without emitting task-log events or touching the
+        shared spinner/interactive state, scoped to the calling thread only,
+        collecting any messages that would have been emitted instead.
+
+        Used by background workers (e.g. the auto grant access worker) that
+        call into commands reporting progress via this UI, so they don't
+        get attributed to - or interrupt - whichever foreground task
+        currently owns the task-log stream, while still keeping a record of
+        what happened. Callers not interested in the messages can simply
+        ignore the yielded list.
+
+        Yields:
+            list[str]: messages emitted during the block, in arrival order.
+        """
+        self._silenced.value = True
+        self._silenced.captured = []
+        try:
+            yield self._silenced.captured
+        finally:
+            self._silenced.captured = None
+            self._silenced.value = False
 
     def print_error(self, msg: str):
         """Display an error message on the command line
@@ -37,6 +71,12 @@ def print_warning(self, msg: str):
         self._print(msg, "warning")
 
     def _print(self, msg: str = "", type: str = "print"):
+        if self._is_silenced:
+            captured = self._captured_messages
+            if captured is not None:
+                captured.append(msg)
+            return
+
         if self.is_interactive:
             self.spinner.write(msg)
         else:
@@ -70,7 +110,10 @@ def interactive(self):
         Yields:
             CLI: Yields the current CLI instance with an interactive session initialized
         """
-        if self.is_interactive:
+        if self._is_silenced:
+            # don't touch the shared spinner/interactive state
+            yield self
+        elif self.is_interactive:
             # if already interactive, do nothing
             yield self
         else:
@@ -97,6 +140,12 @@ def text(self, msg: str = ""):
         # if not self.is_interactive:
         #     self.print(msg)
 
+        if self._is_silenced:
+            captured = self._captured_messages
+            if captured is not None:
+                captured.append(msg)
+            return
+
         self.set_event(
             Event(
                 task_id=self.task_id,
diff --git a/cli/medperf/web_ui/app.py b/cli/medperf/web_ui/app.py
index 3a5f40b92..3b6f1c607 100644
--- a/cli/medperf/web_ui/app.py
+++ b/cli/medperf/web_ui/app.py
@@ -6,8 +6,9 @@
 
 import typer
 from starlette.middleware.base import BaseHTTPMiddleware
-from fastapi import FastAPI, Request
-from fastapi.responses import RedirectResponse
+from fastapi import FastAPI, HTTPException, Request
+from fastapi.exceptions import RequestValidationError
+from fastapi.responses import JSONResponse, RedirectResponse
 from fastapi.staticfiles import StaticFiles
 
 from medperf import config
@@ -91,15 +92,8 @@ def startup_event():
     # List of [schemas.Notification] will appear in the notifications tab
     web_app.state.notifications = []
 
-    # Container auto grant access initial values
-    web_app.state.model_auto_give_access = {
-        "running": False,
-        "worker": None,
-        "benchmark": 0,
-        "model": 0,
-        "emails": "",
-        "interval": 0,
-    }
+    # Dictionary for tracking auto give access status for each model and benchmark combination
+    web_app.state.model_auto_give_access = {}
 
     # Set default UI mode to evaluation on startup, will be updated by NavModeMiddleware on each request based on cookie
     web_app.state.ui_mode = UI_MODE_EVALUATION
@@ -129,6 +123,26 @@ def not_authenticated_exception_handler(
     return RedirectResponse(url=exc.redirect_url)
 
 
+@web_app.exception_handler(HTTPException)
+def http_exception_handler(request: Request, exc: HTTPException):
+    # Reshape into the {"status", "error"} format for consistency with other error responses,
+    # instead of FastAPI's default {"detail": ...} body.
+    return JSONResponse(
+        status_code=exc.status_code,
+        content={"status": "failed", "error": exc.detail},
+    )
+
+
+@web_app.exception_handler(RequestValidationError)
+def validation_exception_handler(request: Request, exc: RequestValidationError):
+    messages = []
+    for error in exc.errors():
+        field = ".".join(str(p) for p in error.get("loc", []) if p != "body")
+        messages.append(f"{field}: {error['msg']}" if field else error["msg"])
+    message = "\n".join(messages) if messages else "Invalid request data."
+    return JSONResponse(status_code=422, content={"status": "failed", "error": message})
+
+
 @web_app.get("/", include_in_schema=False)
 def read_root(request: Request):
     if request.app.state.ui_mode == UI_MODE_TRAINING:
diff --git a/cli/medperf/web_ui/benchmarks/routes.py b/cli/medperf/web_ui/benchmarks/routes.py
index 82a4589aa..8a1499c88 100644
--- a/cli/medperf/web_ui/benchmarks/routes.py
+++ b/cli/medperf/web_ui/benchmarks/routes.py
@@ -1,5 +1,7 @@
 import logging
 
+import anyio
+
 from fastapi import APIRouter, Depends, Form
 from fastapi.responses import HTMLResponse, JSONResponse
 from fastapi import Request
@@ -288,11 +290,17 @@ def update_associations_policy(
     request: Request,
     benchmark_id: int = Form(...),
     dataset_mode: Optional[str] = Form(None),
-    dataset_emails: Optional[str] = Form(None),
     model_mode: Optional[str] = Form(None),
-    model_emails: Optional[str] = Form(None),
     current_user: bool = Depends(check_user_api),
 ):
+    # dataset_emails/model_emails are read from the raw form instead of via
+    # FastAPI's Form(None): an empty-string value there is indistinguishable
+    # from the field being absent, which breaks "explicitly clear the allow
+    # list" (as opposed to "the field wasn't part of this submission at all").
+    form_data = anyio.from_thread.run(lambda: request.form())
+    dataset_emails = form_data.get("dataset_emails")
+    model_emails = form_data.get("model_emails")
+
     initialize_state_task(request, task_name="update_associations_policy")
     return_response = {"status": "", "error": ""}
     try:
@@ -325,7 +333,7 @@ def update_associations_policy(
 def update_committee_members(
     request: Request,
     benchmark_id: int = Form(...),
-    committee_emails: Optional[str] = Form(None),
+    committee_emails: str = Form(""),
     current_user: bool = Depends(check_user_api),
 ):
     initialize_state_task(request, task_name="update_committee_members")
diff --git a/cli/medperf/web_ui/common.py b/cli/medperf/web_ui/common.py
index 6b0dd53d5..9cfa81e12 100644
--- a/cli/medperf/web_ui/common.py
+++ b/cli/medperf/web_ui/common.py
@@ -39,6 +39,7 @@
     "/current_task",
     "/api/running_tasks",
     "/api/stop_task",
+    "/containers/auto_access_logs",
 ]
 
 
diff --git a/cli/medperf/web_ui/containers/routes.py b/cli/medperf/web_ui/containers/routes.py
index 4eb3d10e5..11fc8260d 100644
--- a/cli/medperf/web_ui/containers/routes.py
+++ b/cli/medperf/web_ui/containers/routes.py
@@ -1,5 +1,7 @@
 import logging
 import threading
+from collections import deque
+from datetime import datetime
 from typing import Optional
 
 from fastapi import APIRouter, Depends, Form, Request
@@ -31,6 +33,33 @@
 logger = logging.getLogger(__name__)
 
 
+def _auto_access_key(model_id: int, benchmark_id: int) -> str:
+    return f"{model_id}-{benchmark_id}"
+
+
+def _format_auto_access_log_line(message: str) -> str:
+    timestamp = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+    return f"[{timestamp}] {message}"
+
+
+def _running_auto_access_for_container(
+    model_auto_give_access: dict, container_id: int
+) -> dict:
+    running = {}
+    for key, state in model_auto_give_access.items():
+        try:
+            model_id_str, benchmark_id_str = key.split("-", 1)
+            if int(model_id_str) == container_id:
+                running[int(benchmark_id_str)] = {
+                    "name": state["name"],
+                    "emails": state["emails"],
+                    "interval": state["interval"],
+                }
+        except (ValueError, IndexError):
+            continue
+    return running
+
+
 @router.get("/ui", response_class=HTMLResponse)
 def containers_ui(
     request: Request,
@@ -219,6 +248,14 @@ def container_access_ui(
             if cert_id in certs_mapping:
                 existing_keys[key_id] = certs_mapping[cert_id]
 
+    running_auto_access = _running_auto_access_for_container(
+        request.app.state.model_auto_give_access, container_id
+    )
+    running_benchmarks = [
+        {"id": benchmark_id, "name": state["name"]}
+        for benchmark_id, state in running_auto_access.items()
+    ]
+
     return templates.TemplateResponse(
         "container/container_access.html",
         {
@@ -228,6 +265,8 @@ def container_access_ui(
             "is_owner": is_owner,
             "benchmark_allowed_ids": benchmark_allowed_ids,
             "keys": existing_keys,
+            "running_auto_access": running_auto_access,
+            "running_benchmarks": running_benchmarks,
         },
     )
 
@@ -237,7 +276,7 @@ def grant_access(
     request: Request,
     benchmark_id: int = Form(...),
     model_id: int = Form(...),
-    emails: str = Form(...),
+    emails: str = Form(""),
     current_user: bool = Depends(check_user_api),
 ):
 
@@ -266,19 +305,23 @@ def grant_access(
 
 
 def grant_access_worker(
-    benchmark_id, model_id, emails, interval, stop_event: threading.Event
+    benchmark_id, model_id, emails, interval, stop_event: threading.Event, logs: deque
 ):
     interval_in_seconds = interval * 60
     while not stop_event.is_set():
         try:
-            GrantAccess.run(
-                benchmark_id=benchmark_id,
-                model_id=model_id,
-                emails=emails,
-                approved=True,
-            )
-        except Exception:
-            pass
+            with config.ui.capture() as messages:
+                GrantAccess.run(
+                    benchmark_id=benchmark_id,
+                    model_id=model_id,
+                    approved=True,
+                    allowed_emails=emails,
+                )
+        except Exception as exp:
+            messages.append(f"Error: {exp}")
+            logger.exception(exp)
+        for message in messages:
+            logs.append(_format_auto_access_log_line(message))
         if stop_event.wait(interval_in_seconds):
             break
 
@@ -289,26 +332,36 @@ def start_auto_access(
     benchmark_id: int = Form(...),
     model_id: int = Form(...),
     interval: int = Form(...),
-    emails: str = Form(...),
+    emails: str = Form(""),
     current_user: bool = Depends(check_user_api),
 ):
-    if request.app.state.model_auto_give_access["running"]:
-        bmk = request.app.state.model_auto_give_access["benchmark"]
-        model = request.app.state.model_auto_give_access["model"]
+    model_auto_give_access = request.app.state.model_auto_give_access
+    key = _auto_access_key(model_id, benchmark_id)
+    if key in model_auto_give_access:
         return {
             "status": "failed",
-            "error": f"Auto give access is already running for benchmark: {bmk}, model: {model}",
+            "error": "Auto give access is already running for the selected container and benchmark.",
         }
 
     return_response = {"status": "", "error": ""}
     try:
+        benchmark_name = Benchmark.get(benchmark_id).name
         event = threading.Event()
+        logs = deque(maxlen=config.webui_max_log_messages)
         auto_access_worker = threading.Thread(
             target=grant_access_worker,
-            args=(benchmark_id, model_id, emails, interval, event),
+            args=(benchmark_id, model_id, emails, interval, event, logs),
             daemon=True,
         )
         auto_access_worker.start()
+        model_auto_give_access[key] = {
+            "worker": auto_access_worker,
+            "event": event,
+            "name": benchmark_name,
+            "emails": emails,
+            "interval": interval,
+            "logs": logs,
+        }
         return_response["status"] = "success"
         notification_message = "Successfully started automatic grant access."
     except Exception as exp:
@@ -322,16 +375,6 @@ def start_auto_access(
         return_response=return_response,
         url=f"/containers/ui/display/{model_id}/access",
     )
-
-    request.app.state.model_auto_give_access = {
-        "running": True,
-        "worker": auto_access_worker,
-        "event": event,
-        "benchmark": benchmark_id,
-        "model": model_id,
-        "emails": emails,
-        "interval": interval,
-    }
     return return_response
 
 
@@ -339,9 +382,12 @@ def start_auto_access(
 def stop_auto_access(
     request: Request,
     model_id: int = Form(...),
+    benchmark_id: int = Form(...),
     current_user: bool = Depends(check_user_api),
 ):
-    if not request.app.state.model_auto_give_access["running"]:
+    model_auto_give_access = request.app.state.model_auto_give_access
+    key = _auto_access_key(model_id, benchmark_id)
+    if key not in model_auto_give_access:
         return {
             "status": "failed",
             "error": "Auto give access is not started, nothing to stop.",
@@ -349,8 +395,9 @@ def stop_auto_access(
 
     return_response = {"status": "", "error": ""}
     try:
-        request.app.state.model_auto_give_access["event"].set()
-        request.app.state.model_auto_give_access["worker"].join()
+        model_auto_give_access[key]["event"].set()
+        model_auto_give_access[key]["worker"].join()
+        del model_auto_give_access[key]
         return_response["status"] = "success"
         notification_message = "Successfully stopped automatic grant access."
     except Exception as exp:
@@ -364,18 +411,26 @@ def stop_auto_access(
         return_response=return_response,
         url=f"/containers/ui/display/{model_id}/access",
     )
+    return return_response
 
-    request.app.state.model_auto_give_access = {
-        "running": False,
-        "worker": None,
-        "event": None,
-        "benchmark": 0,
-        "model": 0,
-        "emails": "",
-        "interval": 0,
-    }
 
-    return return_response
+@router.get("/auto_access_logs", response_class=JSONResponse)
+def auto_access_logs(
+    request: Request,
+    model_id: int,
+    benchmark_id: int,
+    current_user: bool = Depends(check_user_api),
+):
+    model_auto_give_access = request.app.state.model_auto_give_access
+    key = _auto_access_key(model_id, benchmark_id)
+    if key not in model_auto_give_access:
+        return {
+            "status": "failed",
+            "error": "Auto give access is not running for the selected container and benchmark.",
+            "logs": [],
+        }
+
+    return {"status": "success", "error": "", "logs": list(model_auto_give_access[key]["logs"])}
 
 
 @router.post("/revoke_user_access", response_class=JSONResponse)
diff --git a/cli/medperf/web_ui/medperf_login.py b/cli/medperf/web_ui/medperf_login.py
index ba448a550..ed9403b45 100644
--- a/cli/medperf/web_ui/medperf_login.py
+++ b/cli/medperf/web_ui/medperf_login.py
@@ -102,6 +102,12 @@ def logout(
     request: Request,
     current_user: bool = Depends(check_user_api),
 ):
+    if request.app.state.model_auto_give_access:
+        return {
+            "status": "failed",
+            "error": "Automatic grant access is currently running. Stop it before logging out.",
+        }
+
     initialize_state_task(request, task_name="medperf_logout")
     return_response = {"status": "", "error": ""}
 
diff --git a/cli/medperf/web_ui/static/js/benchmarks/benchmark_detail.js b/cli/medperf/web_ui/static/js/benchmarks/benchmark_detail.js
index 73919b752..47f0d7a7c 100644
--- a/cli/medperf/web_ui/static/js/benchmarks/benchmark_detail.js
+++ b/cli/medperf/web_ui/static/js/benchmarks/benchmark_detail.js
@@ -48,24 +48,23 @@ function showErrorToast(message) {
     showToast("Validation Error", message, "text-bg-danger");
 }
 
-function checkUpdateAssociationsPolicyForm() {
+function buildAssociationsPolicyConfirmMessage(message) {
     var datasetModeEl = document.getElementById("dataset-auto-approve-mode");
     var modelModeEl = document.getElementById("model-auto-approve-mode");
-    var datasetApproveMode = datasetModeEl ? datasetModeEl.value : "NEVER";
-    var modelApproveMode = modelModeEl ? modelModeEl.value : "NEVER";
-    var isDatasetValid = (datasetApproveMode === "NEVER" || datasetApproveMode === "ALWAYS");
-    var isModelValid = (modelApproveMode === "NEVER" || modelApproveMode === "ALWAYS");
-    if (!isDatasetValid) {
+    var warnings = [];
+    if (datasetModeEl && datasetModeEl.value === "ALLOWLIST") {
         var datasetAllowListArr = getEmailsList(document.getElementById("dataset-allow-list-emails"));
-        if (datasetAllowListArr.length) isDatasetValid = true;
-        else showErrorToast("Make sure that the dataset allow list is not empty");
+        if (!datasetAllowListArr.length) warnings.push("dataset");
     }
-    if (!isModelValid) {
+    if (modelModeEl && modelModeEl.value === "ALLOWLIST") {
         var modelAllowListArr = getEmailsList(document.getElementById("model-allow-list-emails"));
-        if (modelAllowListArr.length) isModelValid = true;
-        else showErrorToast("Make sure that the model allow list is not empty");
+        if (!modelAllowListArr.length) warnings.push("model");
     }
-    return isDatasetValid && isModelValid;
+    if (!warnings.length) return message;
+    var listWord = warnings.length > 1 ? "allow lists are" : "allow list is";
+    return message + " <strong>Note: the " + warnings.join(" and ") + " " + listWord + " empty " +
+        "- no " + warnings.join("/") + " associations will be auto-approved until " +
+        "you add emails.</strong>";
 }
 
 function onUpdateAssociationsPolicySuccess(response) {
@@ -168,7 +167,8 @@ function initBenchmarkDetail() {
 
     var savePolicyBtn = document.getElementById("save-policy-btn");
     if (savePolicyBtn) savePolicyBtn.addEventListener("click", function (e) {
-        if (checkUpdateAssociationsPolicyForm()) showConfirmModal(e.currentTarget, updateAssociationsPolicy, "update benchmark associations policy?");
+        var message = buildAssociationsPolicyConfirmMessage("update benchmark associations policy?");
+        showConfirmModal(e.currentTarget, updateAssociationsPolicy, message);
     });
 
     var saveCommitteeBtn = document.getElementById("save-committee-members-btn");
diff --git a/cli/medperf/web_ui/static/js/common.js b/cli/medperf/web_ui/static/js/common.js
index 5f37e7139..e04e24ae0 100644
--- a/cli/medperf/web_ui/static/js/common.js
+++ b/cli/medperf/web_ui/static/js/common.js
@@ -199,6 +199,7 @@ function showErrorModal(errorTitle, response) {
     var responseError = (response && response.error) || "";
     var responseStatus = (response && response.status) || "";
     var errorText = (responseError + (responseError ? "<br>" : "") + responseStatus).replace(/\n/g, "<br>");
+    if (!errorText) errorText = "Something went wrong. Please try again.";
     var modalBody = "<p id=\"error-text\" class=\"text-lg font-bold text-danger\">" + errorText + "</p><p class=\"text-end mt-3\"><button type=\"button\" class=\"btn btn-xs btn-secondary\" onclick=\"reloadPage();\">Click here to reload</button></p>";
     var modalFooter = "<button type=\"button\" class=\"btn btn-sm btn-danger close-modal-btn\">Hide</button>";
     showModal({ title: errorTitle, body: modalBody, footer: modalFooter });
diff --git a/cli/medperf/web_ui/static/js/containers/container_access.js b/cli/medperf/web_ui/static/js/containers/container_access.js
index 27ef5f4ee..c71d50e35 100644
--- a/cli/medperf/web_ui/static/js/containers/container_access.js
+++ b/cli/medperf/web_ui/static/js/containers/container_access.js
@@ -19,32 +19,123 @@ function createEmailChip(email, inputElement) {
     if (inputElement && inputElement.parentNode) inputElement.parentNode.insertBefore(chip, inputElement);
 }
 
+function clearEmailChips(container) {
+    if (!container) return;
+    container.querySelectorAll(".email-chip").forEach(function (chip) { chip.remove(); });
+}
+
+function setEmailChips(container, emails) {
+    clearEmailChips(container);
+    var inputEl = container ? container.querySelector("input") : null;
+    (emails || []).forEach(function (email) {
+        email = (email || "").trim();
+        if (email) createEmailChip(email, inputEl);
+    });
+}
+
 function parseEmails(element) {
     if (!element || !element.getAttribute) return;
     var raw = element.getAttribute("data-allowed-list") || "[]";
     try {
         var jsonList = JSON.parse(raw);
-        var inputEl = element.querySelector("input");
-        jsonList.forEach(function (email) { createEmailChip(email, inputEl); });
+        setEmailChips(element, jsonList);
     } catch (_) {}
 }
 
+function parseRunningAutoAccess(panel) {
+    if (!panel) return {};
+    try {
+        return JSON.parse(panel.getAttribute("data-running-auto-access") || "{}");
+    } catch (_) {
+        return {};
+    }
+}
+
+function getSelectedBenchmarkId() {
+    var benchmarkEl = document.getElementById("benchmark-auto");
+    return benchmarkEl && benchmarkEl.value ? benchmarkEl.value : "";
+}
+
+function getRunningStateForBenchmark(runningAutoAccess, benchmarkId) {
+    if (!benchmarkId) return null;
+    return runningAutoAccess[benchmarkId] || null;
+}
+
+function parseStoredEmails(emails) {
+    if (!emails) return [];
+    return String(emails).trim().split(/\s+/).filter(Boolean);
+}
+
+function setElementVisible(element, visible) {
+    if (!element) return;
+    element.style.display = visible ? "" : "none";
+    element.classList.toggle("hidden", !visible);
+}
+
+function updateAutoAccessUI() {
+    var panel = document.getElementById("auto-access-panel");
+    var actionsEl = document.getElementById("auto-access-actions");
+    var startBtn = document.getElementById("start-auto-access-btn");
+    var stopBtn = document.getElementById("stop-auto-access-btn");
+    var viewLogsBtn = document.getElementById("view-auto-access-logs-btn");
+    var runningBadge = document.getElementById("running-badge");
+    var intervalEl = document.getElementById("interval-auto");
+    var emailContainer = document.getElementById("allowed-email-list-auto");
+    var emailInput = document.getElementById("email-input-auto");
+    var benchmarkId = getSelectedBenchmarkId();
+    var runningAutoAccess = parseRunningAutoAccess(panel);
+    var runningState = getRunningStateForBenchmark(runningAutoAccess, benchmarkId);
+    var isRunning = Boolean(runningState);
+
+    setElementVisible(startBtn, false);
+    setElementVisible(stopBtn, false);
+    setElementVisible(viewLogsBtn, false);
+    setElementVisible(runningBadge, false);
+
+    if (!benchmarkId) {
+        setElementVisible(actionsEl, false);
+        if (intervalEl) {
+            intervalEl.value = "5";
+            intervalEl.disabled = true;
+        }
+        if (emailInput) emailInput.disabled = true;
+        clearEmailChips(emailContainer);
+        return;
+    }
+
+    setElementVisible(actionsEl, true);
+
+    if (isRunning) {
+        if (intervalEl) {
+            intervalEl.value = runningState.interval || 5;
+            intervalEl.disabled = true;
+        }
+        if (emailInput) emailInput.disabled = true;
+        setEmailChips(emailContainer, parseStoredEmails(runningState.emails));
+        setElementVisible(stopBtn, true);
+        setElementVisible(viewLogsBtn, true);
+        setElementVisible(runningBadge, true);
+    } else {
+        if (intervalEl) {
+            intervalEl.value = "5";
+            intervalEl.disabled = false;
+        }
+        if (emailInput) emailInput.disabled = false;
+        clearEmailChips(emailContainer);
+        setElementVisible(startBtn, true);
+    }
+}
+
 function checkAccessForm() {
-    var allowListArr = getEmailsList(document.getElementById("allowed-email-list"));
     if (!document.getElementById("benchmark") || !document.getElementById("benchmark").value) {
         showErrorToast("Make sure that you've selected a benchmark");
         return false;
     }
-    if (!allowListArr.length) {
-        showErrorToast("Make sure that the email allow list is not empty");
-        return false;
-    }
     return true;
 }
 
 function checkAutoAccessForm() {
-    var allowListArr = getEmailsList(document.getElementById("allowed-email-list-auto"));
-    if (!document.getElementById("benchmark-auto") || !document.getElementById("benchmark-auto").value) {
+    if (!getSelectedBenchmarkId()) {
         showErrorToast("Make sure that you've selected a benchmark");
         return false;
     }
@@ -54,19 +145,22 @@ function checkAutoAccessForm() {
         showErrorToast("Make sure that the time interval is between 5 and 60 (inclusive)");
         return false;
     }
-    if (!allowListArr.length) {
-        showErrorToast("Make sure that the email allow list is not empty");
-        return false;
-    }
     return true;
 }
 
+function emptyAllowListWarning(allowListArr, message) {
+    if (allowListArr.length) return message;
+    return message + " <strong>Note: no emails were added - this will grant access to ALL " +
+        "eligible data owners, with no email filtering.</strong>";
+}
+
 function startAutoGrant(startBtn) {
     disableElements(".card button, .card input, .card select");
+    var panel = document.getElementById("auto-access-panel");
     var allowListArr = getEmailsList(document.getElementById("allowed-email-list-auto"));
     var formData = new FormData();
-    formData.append("benchmark_id", document.getElementById("benchmark-auto").value);
-    formData.append("model_id", startBtn.getAttribute("data-model-id"));
+    formData.append("benchmark_id", getSelectedBenchmarkId());
+    formData.append("model_id", panel ? panel.getAttribute("data-model-id") : "");
     formData.append("interval", document.getElementById("interval-auto").value);
     formData.append("emails", allowListArr.join(" "));
     ajaxRequest("/containers/start_auto_access", "POST", formData, function (response) {
@@ -77,14 +171,33 @@ function startAutoGrant(startBtn) {
 
 function stopAutoGrant(stopBtn) {
     disableElements(".card button, .card input, .card select");
+    var panel = document.getElementById("auto-access-panel");
     var formData = new FormData();
-    formData.append("model_id", stopBtn.getAttribute("data-model-id"));
+    formData.append("model_id", panel ? panel.getAttribute("data-model-id") : "");
+    formData.append("benchmark_id", getSelectedBenchmarkId());
     ajaxRequest("/containers/stop_auto_access", "POST", formData, function (response) {
         if (response && response.status === "success") showReloadModal({ title: "Successfully Stopped Auto Grant Access", seconds: 2 });
         else showErrorModal("Failed to Stop Auto Grant Access", response);
     }, "Failed to stop auto grant access");
 }
 
+function viewAutoAccessLogs() {
+    var panel = document.getElementById("auto-access-panel");
+    var modelId = panel ? panel.getAttribute("data-model-id") : "";
+    var benchmarkId = getSelectedBenchmarkId();
+    var url = "/containers/auto_access_logs?model_id=" + encodeURIComponent(modelId) +
+        "&benchmark_id=" + encodeURIComponent(benchmarkId);
+    ajaxRequest(url, "GET", null, function (response) {
+        var logs = (response && response.logs) || [];
+        var body = logs.length
+            ? "<pre class=\"text-sm overflow-x-auto whitespace-pre-wrap font-mono bg-muted text-ink p-4 rounded-lg max-h-[60vh] overflow-y-auto\">" +
+                logs.map(function (line) { return escapeHtml(cleanMsg(line)); }).join("\n") + "</pre>"
+            : "<p class=\"text-muted-fg\">No logs recorded yet.</p>";
+        var footer = "<button type=\"button\" class=\"btn btn-sm btn-secondary close-modal-btn\">Close</button>";
+        showModal({ title: "Automatic Grant Access Logs", body: body, footer: footer, modalClasses: "max-w-2xl" });
+    }, "Failed to fetch automatic grant access logs");
+}
+
 function showErrorToast(message) {
     showToast("Validation Error", message, "text-bg-danger");
 }
@@ -95,7 +208,6 @@ function isValidEmail(email) {
 
 function init() {
     parseEmails(document.getElementById("allowed-email-list"));
-    parseEmails(document.getElementById("allowed-email-list-auto"));
     document.querySelectorAll(".email-input").forEach(function (input) {
         input.addEventListener("keydown", function (e) {
             if (e.key === "Enter" || e.key === " " || e.key === ",") {
@@ -129,12 +241,25 @@ function init() {
             }
         });
     });
+
+    var benchmarkAutoEl = document.getElementById("benchmark-auto");
+    if (benchmarkAutoEl) benchmarkAutoEl.addEventListener("change", updateAutoAccessUI);
+
     var startBtn = document.getElementById("start-auto-access-btn");
     if (startBtn) startBtn.addEventListener("click", function (e) {
-        if (checkAutoAccessForm()) showConfirmModal(e.currentTarget, startAutoGrant, "start automatic grant access for the selected benchmark?");
+        if (!checkAutoAccessForm()) return;
+        var allowListArr = getEmailsList(document.getElementById("allowed-email-list-auto"));
+        var message = emptyAllowListWarning(allowListArr, "start automatic grant access for the selected benchmark?");
+        showConfirmModal(e.currentTarget, startAutoGrant, message);
     });
     var stopBtn = document.getElementById("stop-auto-access-btn");
-    if (stopBtn) stopBtn.addEventListener("click", function (e) { showConfirmModal(e.currentTarget, stopAutoGrant, "stop automatic grant access?"); });
+    if (stopBtn) stopBtn.addEventListener("click", function (e) {
+        showConfirmModal(e.currentTarget, stopAutoGrant, "stop automatic grant access?");
+    });
+    var viewLogsBtn = document.getElementById("view-auto-access-logs-btn");
+    if (viewLogsBtn) viewLogsBtn.addEventListener("click", viewAutoAccessLogs);
+
+    updateAutoAccessUI();
 }
 if (document.readyState === "loading") document.addEventListener("DOMContentLoaded", init);
 else init();
diff --git a/cli/medperf/web_ui/templates/container/container_access.html b/cli/medperf/web_ui/templates/container/container_access.html
index e63a8290f..1f460a036 100644
--- a/cli/medperf/web_ui/templates/container/container_access.html
+++ b/cli/medperf/web_ui/templates/container/container_access.html
@@ -3,14 +3,6 @@
 {% block title %}Manage Container Access{% endblock %}
 
 {% block detail_panel %}
-{% set auto_access_running = request.app.state.model_auto_give_access["running"] %}
-{% set auto_access_for_this_model = auto_access_running and request.app.state.model_auto_give_access["model"] == entity.id %}
-{% if auto_access_for_this_model %}
-    {% set selected_benchmark = request.app.state.model_auto_give_access["benchmark"] %}
-    {% set selected_interval = request.app.state.model_auto_give_access["interval"] %}
-    {% set selected_emails = request.app.state.model_auto_give_access["emails"] %}
-{% endif %}
-
 <div class="w-full max-w-[1440px] mx-auto px-4 sm:px-6 lg:px-8">
     <a href="/containers/ui/display/{{ entity.id }}" class="text-muted-fg hover:underline">← Back to Container Details</a>
     <h1 class="text-2xl font-bold mt-4 text-ink">Manage Access | {{ entity_name }}</h1>
@@ -44,9 +36,18 @@ <h3 class="text-lg font-bold text-ink mb-2">Grant Access</h3>
         </div>
     </form>
 
-    <div class="bg-card dark:bg-card rounded-2xl p-6 shadow-lg border-2 border-border-brand dark:border-border mb-6">
+    <div id="auto-access-panel" class="bg-card dark:bg-card rounded-2xl p-6 shadow-lg border-2 border-border-brand dark:border-border mb-6" data-model-id="{{ entity.id }}" data-running-auto-access='{{ running_auto_access | tojson }}'>
         <h3 class="text-lg font-bold text-ink mb-2">Automatic Access</h3>
         <p class="text-muted-fg mb-4">Automatically grant access to new data owners associated to a benchmark.</p>
+        {% if running_benchmarks %}
+        <p class="text-sm text-muted-fg mb-4">Currently running for:
+            {% for benchmark in running_benchmarks %}
+            <a href="/benchmarks/ui/display/{{ benchmark.id }}" class="font-bold text-brand-accent hover:underline">{{ benchmark.name }}</a>{% if not loop.last %}, {% endif %}
+            {% endfor %}
+        </p>
+        <p class="text-sm text-muted-fg mb-4">Note: select the benchmark below to stop it or view its logs.</p>
+        {% endif %}
+
         <div class="grid grid-cols-1 md:grid-cols-2 gap-4 mb-4">
             <div>
                 <label for="benchmark-auto" class="block text-secondary font-semibold mb-2">Benchmark</label>
@@ -55,28 +56,26 @@ <h3 class="text-lg font-bold text-ink mb-2">Automatic Access</h3>
                 {% set input_id = "benchmark-auto" %}
                 {% set placeholder = "Search approved benchmarks…" %}
                 {% set allowed_ids = benchmark_allowed_ids %}
-                {% set selected_value = selected_benchmark if auto_access_for_this_model else '' %}
-                {% set disabled = auto_access_for_this_model %}
+                {% set required = false %}
+                {% set disabled = task_running %}
                 {% include "partials/searchable_select.html" %}
             </div>
             <div>
                 <label for="interval-auto" class="block text-secondary font-semibold mb-2">Interval (minutes)</label>
-                <input type="number" id="interval-auto" class="w-full px-4 py-3 rounded-xl border-2 border-border-strong dark:border-border dark:bg-card-muted dark:text-ink focus:outline-none focus:border-brand-accent" min="5" max="60" value="{% if auto_access_for_this_model %}{{ selected_interval }}{% else %}5{% endif %}" {% if auto_access_for_this_model %}disabled{% endif %}>
+                <input type="number" id="interval-auto" class="w-full px-4 py-3 rounded-xl border-2 border-border-strong dark:border-border dark:bg-card-muted dark:text-ink focus:outline-none focus:border-brand-accent" min="5" max="60" value="5" disabled>
             </div>
         </div>
         <div class="mb-4" id="email-chip-container-auto">
             <label for="email-input-auto" class="block text-secondary font-semibold mb-2">Allow list emails</label>
-            <div class="email-container" id="allowed-email-list-auto" data-allowed-list='{% if auto_access_for_this_model %}{{ selected_emails }}{% else %}{}{% endif %}'>
-                <input type="text" name="emails" id="email-input-auto" class="email-input w-full px-4 py-3 rounded-xl border-2 border-border-strong dark:border-border dark:bg-card-muted dark:text-ink mt-2 focus:outline-none focus:border-brand-accent" placeholder="Type email then hit Enter, comma, or space" {% if task_running %}disabled{% endif %}>
+            <div class="email-container" id="allowed-email-list-auto" data-allowed-list='[]'>
+                <input type="text" name="emails" id="email-input-auto" class="email-input w-full px-4 py-3 rounded-xl border-2 border-border-strong dark:border-border dark:bg-card-muted dark:text-ink mt-2 focus:outline-none focus:border-brand-accent" placeholder="Type email then hit Enter, comma, or space" disabled>
             </div>
         </div>
-        <div class="text-center">
-            {% if auto_access_for_this_model %}
-            <button type="button" class="btn btn-danger step-btn" id="stop-auto-access-btn" data-model-id="{{ entity.id }}">Stop Automatic Access</button>
-            <span id="running-badge" class="px-3 py-1 rounded-full text-sm font-bold bg-success-muted text-success ml-3 align-middle inline-block">Running</span>
-            {% else %}
-            <button type="button" class="btn btn-info step-btn" id="start-auto-access-btn" data-model-id="{{ entity.id }}">Start Automatic Access</button>
-            {% endif %}
+        <div id="auto-access-actions" class="text-center hidden">
+            <button type="button" class="btn btn-info step-btn hidden" id="start-auto-access-btn" {% if task_running %}disabled{% endif %}>Start Automatic Access</button>
+            <button type="button" class="btn btn-danger step-btn hidden" id="stop-auto-access-btn" {% if task_running %}disabled{% endif %}>Stop Automatic Access</button>
+            <button type="button" class="btn btn-secondary step-btn hidden" id="view-auto-access-logs-btn">View Logs</button>
+            <span id="running-badge" class="hidden px-3 py-1 rounded-full text-sm font-bold bg-success-muted text-success ml-3 align-middle inline-block">Running</span>
         </div>
     </div>