Which AWS service should the company use to build this capability?

Explanations:

Amazon SageMaker is a fully managed service that enables developers to build, train, and deploy machine learning models, requiring expertise in ML. It is not suitable for users without ML knowledge who need to identify inappropriate photos.

Amazon Textract is a service that automatically extracts text and data from scanned documents. It is not designed for image analysis or detecting inappropriate content in photos.

Amazon Rekognition is a deep learning-based image and video analysis service that can detect inappropriate content in images. It provides pre-trained models for various tasks, including explicit content detection, making it suitable for users without ML expertise.

Amazon Comprehend is a natural language processing (NLP) service that analyzes text to find insights and relationships. It does not handle image content and is therefore not applicable for detecting inappropriate photos.