Engineering students at KAUST tackled AI bias with inclusive solutions
![](https://cdn.prod.website-files.com/64a73265ea06a02bc3eab456/64f5a38681a3228af035f8a1_DpULtWHW0AADi48.jpeg)
The students participated in challenges revolving around discovering AI bias in the "hack AI for All" event
Saudi engineering students recently convened at King Abdullah University of Science and Technology (KAUST) from August 20 to 24 for the "hack AI for All" hackathon. They aimed to address biases embedded in artificial intelligence (AI) software. These students, enrolled in an AI summer course at both KAUST and Oxford University, engaged in challenges centered around bias detection, language enhancement, and safety improvement. This event marked the second phase of KAUST's "Dear AI" campaign, launched to raise awareness about AI neural network training deficiencies and promote a more inclusive world through innovative solutions.
The hackathon addressed the pressing issue of AI biases stemming from societal prejudices present during training. As AI's influence grows, it's imperative for search results to accurately reflect diverse user populations. For instance, women's underrepresentation in AI search outputs underscores the importance of accurate AI representation. The Ministry of Communications and Information Technology (MCIT), Saudi Data and Artificial Intelligence Authority (SDAIA), and global marketing agency VMLY&R collaborated with KAUST to organize this transformative event. The winning team, Hunayn, excelled in refining AI models for Arabic language processing, embracing the rich semantics of Arabic. By participating, these emerging undergraduates gained a broader understanding of AI's landscape, fostering newfound confidence in their capacity to shape its future.