If my program will work ONLY with Unicode encoded files, can I consider that if BOM is not presented, encoding is UTF-8 without BOM?

My question namely is "Are UTF-16 and UTF-32 encoded files considered correct if BOM is not presented?"

I worked on it and found that BOM is optional for UTF-16 and UTF-32, it`s considered good practice to ask the user about encoding if BOM is not presented. I got information from this article.