We propose a method for detecting tampering in recordings using a blind watermarking detection method and audio fingerprinting. In the proposed method, fingerprints extracted from audio recordings are embedded into audio recordings as watermarks for tamper detection. At the decoder, this is matched with the fingerprint extracted from the audio recording itself. To achieve blind watermark detection, a Wiener filter is used to estimate the clean speech, subtracted from the recorded audio, and the embedded watermark is extracted from the residual signal. The vector distance between the fingerprints obtained from the watermark and the fingerprints obtained from the recorded audio is calculated. If the distance is greater than a predetermined threshold, it is judged to have been tampered with, and if it is less, the authenticity of the recording can be proven. We used recordings degraded with reverberation and additive noise to evaluate the detection accuracy of substitution tampering, and the detection accuracy of 99.87% was shown.