New Delhi, India – What if Michael had died instead of Sonny in The Godfather? Or if Rose had shared the debris plank, and Jack hadn't been left to freeze in the Atlantic in Titanic*? Eros International, one of India's largest production houses, with more than 4,000 films in its catalogue, has decided to explore this sort of what-if scenario. It has re-released one of its major hits, Raanjhanaa, a 2013 romantic drama, in cinemas – but has used artificial intelligence (AI) to change its tragic end, in which the male lead dies. In the AI-altered version, Kundan (played by popular actor Dhanush), a Hindu man who has a doomed romance with a Muslim woman, lives.

Deep neural networks trained on biased data often inadvertently learn unintended inference rules, particularly when labels are strongly correlated with biased features. Existing bias mitigation methods typically involve either a) predefining bias types and enforcing them as prior knowledge or b) reweighting training samples to emphasize bias-conflicting samples over bias-aligned samples. However, both strategies address bias indirectly in the feature or sample space, with no control over learned weights, making it difficult to control the bias propagation across different layers. Based on this observation, we introduce a novel approach to address bias directly in the model's parameter space, preventing its propagation across layers. Our method involves training two models: a bias model for biased features and a debias model for unbiased details, guided by the bias model. We enforce dissimilarity in the debias model's later layers and similarity in its initial layers with the bias model, ensuring it learns unbiased low-level features without adopting biased high-level abstractions. By incorporating this explicit constraint during training, our approach shows enhanced classification accuracy and debiasing effectiveness across various synthetic and real-world datasets of different sizes. Moreover, the proposed method demonstrates robustness across different bias types and percentages of biased samples in the training data. The code is available at: https://visdomlab.github.io/CosFairNet/