It's an open standard because the concepts and reference implementation are free and open source even if the PDF is paywalled. Realistically you could just pirate the PDF and write a jpeg xl encoder/decoder and your code wouldn't be infringing on any patents.