AI's Inclusion Gap: CMU-Q’s expert in Arabic AI calls for research that "reflects our culture and identity"

The future of artificial intelligence in the Middle East hinges on solving its "inclusion" problem, according to Houda Bouamor, a leading expert in Arabic Natural Language Processing (NLP) and an associate teaching professor at Carnegie Mellon University in Qatar (CMU-Q).

Most AI models are trained on formal, literary Arabic, which leaves them unable to understand the region's many spoken dialects. Furthermore, Arabic’s complex morphology creates a level of nuance that English-centric models are not designed to handle.

Boumor argues this is not just a technical issue: "It’s about inclusion and making sure our societies benefit from AI in ways that fit our needs."

The Arabic Language Gap

Even with hundreds of millions of speakers, Arabic is considered a low-resource language in the AI world. This creates a significant disparity in how technology serves the region.

"For everyday dialects — like those spoken here in Qatar — the gap is even bigger," she noted. "This means many tools don’t fully understand how we communicate in real life.”

The impact of this gap is far-reaching. Bouamor explains that solving it is essential for improving education by "supporting children in their own language" and building critical accessibility tools, such as speech recognition for people with disabilities.

A broader regional vision

This linguistic gap is part of a larger set of regional challenges that Bouamor believes AI is uniquely positioned to address.

She identifies multimodality—the integration of speech, images, and video—as an important enabler for a region with rich oral traditions and burgeoning media industries.

Furthermore, she stresses the need for AI for social good, directing research to solve specific Middle East and North Africa (MENA) challenges in healthcare, climate sustainability, and crisis management.

However, this technical growth, she warns, must be paired with ethical considerations.

“Growth should not only be technical,” she said. “We need governance frameworks that ensure AI reflects our cultural values, addresses bias, and is trusted by society. This is a key area for research — and it is particularly important in regions where diverse voices must be represented.”

A new generation of specialists

To meet these complex regional challenges, Carnegie Mellon University in Qatar has launched a new Bachelor of Science in Artificial Intelligence (BSAI). The new program will augment the AI focus in the campus’ existing programs in biological sciences, business administration, computer science and information systems.

Bouamor describes the new degree as a catalyst to creating a new, collaborative ecosystem of local talent.

Within CMU-Q, she looks forward to a powerful synergy between those who build AI and those who apply it: “The BSAI will create specialists who can build advanced models, while students, faculty and researchers in the other programs can guide new areas for discipline-specific applications and responsible, ethical deployment."

This dual focus on capability and conscience is already a core part of the university's approach, and Bouamor looks forward to expanding this as the BSAI program is rolled out: “Our goal is to help create a self-sustaining innovation ecosystem in Qatar, and prepare graduates who can develop cutting-edge AI that responsibly serves people and society."