<p>An image encryption algorithm enhances the security of an image by transforming its content into a scrambled format. This process employs mathematical techniques, including chaos theory, permutation, and diffusion, to prevent unauthorized access or tampering. To improve the security of images during transmission, we introduce an efficient and innovative image encryption algorithm based on a time-delay predator–prey model, integrating principles from ecological dynamics into cryptography. By appropriately selecting parameters of the proposed predator–prey model, chaotic phenomena are generated. We use these chaotic sequences for image encryption and combine them with the Arnold scrambling algorithm to rearrange pixel positions within the image. Additionally, a diffusion algorithm is employed to alter pixel values, achieving the encryption effect. Various experimental analyses, such as initial value sensitivity, histogram analysis, adjacent pixel correlation, and overall robustness evaluations, are conducted. The computational experiments indicate that the chaotic sequence generated by the proposed predator–prey model can effectively implement image encryption with a favorable encryption effect. We present the results of the model on the best Number of Pixel Change Rate (NPCR) and Unified Average Change Intensity (UACI) scores for various selected test images. In the end, the average NPCR and average UACI are 99.6067% and 33.4687% respectively. We also calculated the value of the information entropy indicator, which reaches 7.9986. Through these empirical validations, the numerical results highlight the robustness and viability of using chaotic sequences in the context of image encryption, offering promising prospects for enhancing data security in digital transmission systems.</p>