Introducing zkDL++, a novel framework designed for provable AI. Leveraging zkDL++, weaddress a key challenge in generative AI watermarking—maintaining privacy whileensuring provability. By enhancing the watermarking system developed by Meta, zkDL++solves the problem of needing to keep watermark extractors private to avoid attacks,offering a more secure solution. Beyond watermarking, zkDL++ proves the integrity of anydeep neural network (DNN) with high efficiency. In this post, we outline our approach,evaluate its performance, and propose avenues for further optimization.