A new method called Orthogonal Concept Erasure (OCE) improves the efficiency of concept erasure in diffusion models by using multiplicative parameter updates, allowing for precise erasure of multiple concepts without sacrificing generative capacity. This advancement could enhance the deployment of AI models in sensitive applications by effectively mitigating undesired content.