Why the proposed Artificial Intelligence Regulation does not deliver on the promise to protect individuals from harm