Abstract: Digital audio watermarking technology is a technique for embedding specific identification information into digital audio content, which can be used for copyright protection or content verification. To meet the needs of both copyright protection and content verification, a quantization method based on confidential features was proposed, and a dual watermark algorithm for digital audio based on this feature was constructed. By embedding the watermark in the confidential feature, the security of the watermark system was effectively improved. Moreover, the algorithm could achieve both copyright protection and audio content verification, enhancing its practicality. The experimental results indicated that the algorithm had good imperceptibility, robustness, security and detection ability against malicious attacks.