diff --git a/arm_compute/runtime/ITensorAllocator.h b/arm_compute/runtime/ITensorAllocator.h
index 1aed08fe33..6041f18e4b 100644
--- a/arm_compute/runtime/ITensorAllocator.h
+++ b/arm_compute/runtime/ITensorAllocator.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2021, 2024-2025 Arm Limited.
+ * Copyright (c) 2016-2021, 2024-2026 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -110,10 +110,26 @@ class ITensorAllocator
     /** Interface to be implemented by the child class to unlock the memory allocation after the CPU is done accessing it. */
     virtual void unlock() = 0;
 
+    /** Returns whether allocator currently owns the TensorInfo metadata object. */
+    bool owns_info() const;
+    /** Returns whether the allocator has imported external backing memory. */
+    bool is_imported() const;
+    /** Track import state for derived allocators when ownership of backing memory changes. */
+    void set_imported(bool imported);
+    /** For external TensorInfo metadata, import state determines resizable behavior from allocator perspective. */
+    bool allocator_considers_resizable() const;
+    /** Set the resizable flag on the owned TensorInfo only if the allocator owns it.
+     *
+     * @return True if the TensorInfo was mutated, false otherwise.
+     */
+    bool set_resizable_if_info_owned(bool is_resizable);
+
 private:
     TensorInfo  _info_owned{};           /**< Tensor's metadata. */
     TensorInfo *_info_external{nullptr}; /**< External Tensor's metadata */
     size_t      _alignment{};            /**< Tensor's alignment in bytes */
+    bool        _owns_info{true};        /**< True when allocator owns metadata; false for soft_init(). */
+    bool        _is_imported{false};     /**< True when memory was imported instead of allocated by allocator. */
 };
 } // namespace arm_compute
 #endif // ACL_ARM_COMPUTE_RUNTIME_ITENSORALLOCATOR_H
diff --git a/src/runtime/CL/CLTensorAllocator.cpp b/src/runtime/CL/CLTensorAllocator.cpp
index b109288bfb..c3b15dd487 100644
--- a/src/runtime/CL/CLTensorAllocator.cpp
+++ b/src/runtime/CL/CLTensorAllocator.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2021, 2024 Arm Limited.
+ * Copyright (c) 2016-2021, 2024, 2026 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -157,7 +157,8 @@ void CLTensorAllocator::allocate()
     }
 
     // Lock allocator
-    info().set_is_resizable(false);
+    set_imported(false);
+    set_resizable_if_info_owned(false);
 }
 
 void CLTensorAllocator::free()
@@ -165,7 +166,8 @@ void CLTensorAllocator::free()
     _mapping = nullptr;
     _memory.set_region(nullptr);
     clear_quantization_arrays(_scale, _offset);
-    info().set_is_resizable(true);
+    set_imported(false);
+    set_resizable_if_info_owned(true);
 }
 
 bool CLTensorAllocator::is_allocated() const
@@ -182,7 +184,8 @@ Status CLTensorAllocator::import_memory(cl::Buffer buffer)
 
     _memory.set_owned_region(std::make_unique<CLBufferMemoryRegion>(buffer));
 
-    info().set_is_resizable(false);
+    set_imported(true);
+    set_resizable_if_info_owned(false);
     return Status{};
 }
 
diff --git a/src/runtime/ITensorAllocator.cpp b/src/runtime/ITensorAllocator.cpp
index fe3d2804cb..f3a66a71f6 100644
--- a/src/runtime/ITensorAllocator.cpp
+++ b/src/runtime/ITensorAllocator.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2021 Arm Limited.
+ * Copyright (c) 2016-2021, 2026 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -35,12 +35,16 @@ void ITensorAllocator::init(const TensorInfo &input, size_t alignment)
     _info_owned    = input;
     _info_external = nullptr;
     _alignment     = alignment;
+    _owns_info     = true;
+    _is_imported   = false;
 }
 
 void ITensorAllocator::soft_init(TensorInfo &input, size_t alignment)
 {
     _info_external = &input;
     _alignment     = alignment;
+    _owns_info     = false;
+    _is_imported   = false;
 }
 
 TensorInfo &ITensorAllocator::info()
@@ -57,3 +61,38 @@ size_t ITensorAllocator::alignment() const
 {
     return _alignment;
 }
+
+bool ITensorAllocator::owns_info() const
+{
+    return _owns_info;
+}
+
+bool ITensorAllocator::is_imported() const
+{
+    return _is_imported;
+}
+
+void ITensorAllocator::set_imported(bool imported)
+{
+    _is_imported = imported;
+}
+
+bool ITensorAllocator::allocator_considers_resizable() const
+{
+    if (_owns_info)
+    {
+        const TensorInfo *info_ptr = (_info_external != nullptr) ? _info_external : &_info_owned;
+        return (info_ptr != nullptr) ? info_ptr->is_resizable() : true;
+    }
+    return !_is_imported;
+}
+
+bool ITensorAllocator::set_resizable_if_info_owned(bool is_resizable)
+{
+    if (_owns_info)
+    {
+        info().set_is_resizable(is_resizable);
+        return true;
+    }
+    return false;
+}
diff --git a/src/runtime/TensorAllocator.cpp b/src/runtime/TensorAllocator.cpp
index b803f77522..82ad5ee04b 100644
--- a/src/runtime/TensorAllocator.cpp
+++ b/src/runtime/TensorAllocator.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2016-2020, 2024 Arm Limited.
+ * Copyright (c) 2016-2020, 2024, 2026 Arm Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -71,7 +71,7 @@ TensorAllocator::TensorAllocator(IMemoryManageable *owner) : _owner(owner), _ass
 
 TensorAllocator::~TensorAllocator()
 {
-    info().set_is_resizable(true);
+    set_resizable_if_info_owned(true);
 }
 
 TensorAllocator::TensorAllocator(TensorAllocator &&o) noexcept
@@ -142,13 +142,15 @@ void TensorAllocator::allocate()
     {
         _associated_memory_group->finalize_memory(_owner, _memory, info().total_size(), alignment_to_use);
     }
-    info().set_is_resizable(false);
+    set_imported(false);
+    set_resizable_if_info_owned(false);
 }
 
 void TensorAllocator::free()
 {
     _memory.set_region(nullptr);
-    info().set_is_resizable(true);
+    set_imported(false);
+    set_resizable_if_info_owned(true);
 }
 
 bool TensorAllocator::is_allocated() const
@@ -163,7 +165,8 @@ Status TensorAllocator::import_memory(void *memory)
     ARM_COMPUTE_RETURN_ERROR_ON(alignment() != 0 && !arm_compute::utility::check_aligned(memory, alignment()));
 
     _memory.set_owned_region(std::make_unique<MemoryRegion>(memory, info().total_size()));
-    info().set_is_resizable(false);
+    set_imported(true);
+    set_resizable_if_info_owned(false);
 
     return Status{};
 }
diff --git a/tests/validation/UNIT/shared_import_memory_test.cpp b/tests/validation/UNIT/shared_import_memory_test.cpp
new file mode 100644
index 0000000000..457534d384
--- /dev/null
+++ b/tests/validation/UNIT/shared_import_memory_test.cpp
@@ -0,0 +1,75 @@
+/*
+ * Copyright (c) 2026 Arm Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Coordinates.h"
+#include "arm_compute/core/TensorInfo.h"
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/runtime/Tensor.h"
+
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/validation/Validation.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+TEST_SUITE(UNIT)
+TEST_SUITE(TensorInfo)
+
+TEST_CASE(ImportMemoryDoesNotMutateExternalInfo, framework::DatasetMode::ALL)
+{
+    TensorInfo out_info(TensorShape(16U, 4U, 4U), 1, DataType::F32, DataLayout::NHWC);
+    out_info.set_is_resizable(true);
+
+    Tensor out_tensor;
+    out_tensor.allocator()->init(out_info);
+    out_tensor.allocator()->allocate();
+
+    // Simulate a shared TensorInfo used as a view wrapper.
+    TensorInfo shared_info(out_info);
+    shared_info.set_is_resizable(true);
+
+    Tensor view_tensor;
+    view_tensor.allocator()->soft_init(shared_info);
+
+    // Ensure it's still resizable before import.
+    ARM_COMPUTE_EXPECT(shared_info.is_resizable(), framework::LogLevel::ERRORS);
+
+    // Import memory into the view tensor.
+    ARM_COMPUTE_ASSERT(bool(view_tensor.allocator()->import_memory(out_tensor.buffer())));
+
+    // Regression assert: import_memory must NOT mutate the caller-owned shared_info.
+    ARM_COMPUTE_EXPECT(shared_info.is_resizable(), framework::LogLevel::ERRORS);
+
+    // extend_padding should succeed (not throw).
+    ARM_COMPUTE_EXPECT_NO_THROW(shared_info.extend_padding(PaddingSize(1, 1, 1, 1)), framework::LogLevel::ERRORS);
+}
+
+TEST_SUITE_END() // TensorInfo
+TEST_SUITE_END() // UNIT
+} // namespace validation
+} // namespace test
+} // namespace arm_compute