diff --git a/cpp/src/interop/detail/arrow_allocator.cpp b/cpp/src/interop/detail/arrow_allocator.cpp
index 54fa50d03cb..cb67c893573 100644
--- a/cpp/src/interop/detail/arrow_allocator.cpp
+++ b/cpp/src/interop/detail/arrow_allocator.cpp
@@ -23,7 +23,7 @@ std::unique_ptr<arrow::Buffer> allocate_arrow_buffer(const int64_t size, arrow::
 {
   /*
   nvcc 11.0 generates Internal Compiler Error during codegen when arrow::AllocateBuffer
-  and `ValueOrDie` or are used inside a CUDA compilation unit.
+  and `ValueOrDie` are used inside a CUDA compilation unit.
 
   To work around this issue we compile an allocation shim in C++ and use
   that from our cuda sources