diff --git a/src/core/CMakeLists.txt b/src/core/CMakeLists.txt
index 5469e02a9ea..f14298f15eb 100644
--- a/src/core/CMakeLists.txt
+++ b/src/core/CMakeLists.txt
@@ -31,10 +31,6 @@ set_property(SOURCE ${MIXED_SRC}
         $<TARGET_PROPERTY:inference_engine_obj,SOURCE_DIR>/src
         $<TARGET_PROPERTY:inference_engine_plugin_api,INTERFACE_INCLUDE_DIRECTORIES>)
 
-
-if(CMAKE_CXX_COMPILER_ID STREQUAL "MSVC")
-    ie_add_compiler_flags(/wd4244)
-endif()
 # Create named folders for the sources within the .vcproj
 # Empty name lists them directly under the .vcproj
 
@@ -112,7 +108,6 @@ if(CMAKE_CXX_COMPILER_ID STREQUAL "MSVC")
     endif()
 
     target_link_options(ngraph_obj ${link_type} "/IGNORE:4217,4286")
-    ie_add_compiler_flags(/wd4267)
 endif()
 
 # some sources are located in ngraph, while headers are in inference_engine_transformations
diff --git a/src/core/reference/include/ngraph/runtime/reference/gather.hpp b/src/core/reference/include/ngraph/runtime/reference/gather.hpp
index 95cdc73c83e..82a7fb50188 100644
--- a/src/core/reference/include/ngraph/runtime/reference/gather.hpp
+++ b/src/core/reference/include/ngraph/runtime/reference/gather.hpp
@@ -34,7 +34,7 @@ void gather(const T* const data,
     int64_t axis_size = data_shape[axis];
     int64_t data_offset, out_offset, idx;
     // for out of bound indices is filled with zeros
-    std::fill(out, out + shape_size(out_shape), 0);
+    std::fill(out, out + shape_size(out_shape), T{0});
 
     for (int64_t batch = 0; batch < batch_size; batch++)
         for (int64_t outer_idx = 0; outer_idx < outer_size; outer_idx++) {
diff --git a/src/core/reference/include/ngraph/runtime/reference/grid_sample.hpp b/src/core/reference/include/ngraph/runtime/reference/grid_sample.hpp
index f2662d56173..ed3afda4601 100644
--- a/src/core/reference/include/ngraph/runtime/reference/grid_sample.hpp
+++ b/src/core/reference/include/ngraph/runtime/reference/grid_sample.hpp
@@ -133,10 +133,11 @@ DATA_ET bilinear(const DATA_ET* data,
     const auto x_topleft = std::floor(x_d);
     const auto dy = y_d - y_topleft;
     const auto dx = x_d - x_topleft;
-    const auto v00 = get_padded(data, data_shape, n, c, y_topleft, x_topleft);
-    const auto v01 = get_padded(data, data_shape, n, c, y_topleft, x_topleft + 1);
-    const auto v10 = get_padded(data, data_shape, n, c, y_topleft + 1, x_topleft);
-    const auto v11 = get_padded(data, data_shape, n, c, y_topleft + 1, x_topleft + 1);
+    const auto v00 = get_padded(data, data_shape, n, c, static_cast<long>(y_topleft), static_cast<long>(x_topleft));
+    const auto v01 = get_padded(data, data_shape, n, c, static_cast<long>(y_topleft), static_cast<long>(x_topleft + 1));
+    const auto v10 = get_padded(data, data_shape, n, c, static_cast<long>(y_topleft + 1), static_cast<long>(x_topleft));
+    const auto v11 =
+        get_padded(data, data_shape, n, c, static_cast<long>(y_topleft + 1), static_cast<long>(x_topleft + 1));
 
     const auto q0 = (1 - dx) * v00 + dx * v01;
     const auto q1 = (1 - dx) * v10 + dx * v11;
@@ -204,7 +205,13 @@ DATA_ET bicubic(const DATA_ET* data,
     const auto x_topleft = std::floor(x_d);
     const auto dy = y_d - y_topleft;
     const auto dx = x_d - x_topleft;
-    const auto s = gather_4x4(data, data_shape, n, c, y_topleft - 1, x_topleft - 1, get_padded);
+    const auto s = gather_4x4(data,
+                              data_shape,
+                              n,
+                              c,
+                              static_cast<long>(y_topleft - 1),
+                              static_cast<long>(x_topleft - 1),
+                              get_padded);
 
     const auto cy = cubic_coeffs(dy);
     const auto cx = cubic_coeffs(dx);
diff --git a/src/core/shape_inference/include/space_to_batch_shape_inference.hpp b/src/core/shape_inference/include/space_to_batch_shape_inference.hpp
index 7cc04ec0ac3..9b239e00fc5 100644
--- a/src/core/shape_inference/include/space_to_batch_shape_inference.hpp
+++ b/src/core/shape_inference/include/space_to_batch_shape_inference.hpp
@@ -59,7 +59,7 @@ std::vector<TShape> shape_infer(const SpaceToBatch* op,
 
         auto blocks = get_input_const_data_as<TShape, int64_t>(op, 1, constant_data);
         if (blocks) {
-            TVal block_prod = std::accumulate(begin(*blocks), end(*blocks), 1, std::multiplies<int64_t>());
+            TVal block_prod = std::accumulate(begin(*blocks), end(*blocks), int64_t(1), std::multiplies<int64_t>());
             out_shape.push_back(data_shape[0] * block_prod);
         } else {
             out_shape.emplace_back(dim::inf_bound);
diff --git a/src/core/shape_inference/include/tile_shape_inference.hpp b/src/core/shape_inference/include/tile_shape_inference.hpp
index abc220b999c..465e4a42397 100644
--- a/src/core/shape_inference/include/tile_shape_inference.hpp
+++ b/src/core/shape_inference/include/tile_shape_inference.hpp
@@ -11,6 +11,15 @@ namespace ov {
 namespace op {
 namespace v0 {
 
+template <class T>
+struct NegativeToZero {
+    NegativeToZero() = default;
+    template <class U>
+    T operator()(const U u) const {
+        return static_cast<T>(std::max<U>(0, ov::util::InTypeRange<U>()(u)));
+    }
+};
+
 template <class T>
 std::vector<T> shape_infer(const Tile* op,
                            const std::vector<T>& input_shapes,
@@ -28,9 +37,7 @@ std::vector<T> shape_infer(const Tile* op,
     T output_shape;
 
     // Get repeats and pre process values
-    auto negative_repeats_to_zero = [](const TDimValue v) -> TDimValue {
-        return std::max<TDimValue>(0, ov::util::InTypeRange<TDimValue>()(v));
-    };
+    constexpr auto negative_repeats_to_zero = NegativeToZero<TDimValue>();
 
     auto repeats = get_input_const_data_as_shape<T>(op, 1, constant_data, negative_repeats_to_zero);