How AI is revolutionising life for the visually impaired in the Arab world

For years, Saudi media professional Khalid Al-Harbi waited anxiously for friends to describe the images that arrived on his WhatsApp or explain the infographics that government agencies shared on social media.

As a blind person, he depended entirely on others to understand visual content or read PDF documents for his work.

Today, Khalid smiles as he describes the moment of "liberation" that artificial intelligence brought him.

"Now I can read everything myself – photos, documents," he says, his voice carrying a mix of wonder and relief at the independence these tools have provided.

Khalid's experience reflects a broader trend across the Arab world, where an estimated five million blind individuals and 40.5 million people with visual impairments are discovering unprecedented autonomy through AI technology.

Yet this digital revolution comes with significant obstacles that highlight the gap between technological promise and regional reality.

According to World Health Organization estimates for the Middle East and North Africa region, most cases of visual impairment stem from preventable or treatable causes such as cataracts and uncorrected refractive errors, alongside more serious conditions like glaucoma and diabetic retinopathy.

In Saudi Arabia alone, visual impairment affects an estimated 0.6-0.7% of the population, equivalent to approximately 160,000 blind individuals and 500,000 people with visual impairments.

Amjad Saad Al-Mutairi, a recent King Saud University graduate working in photography, spent her entire academic career relying on her close friend Samar to read textbooks aloud.

The arrangement worked, but it meant her studies moved at someone else's pace, her learning constrained by another person's availability.

But then, AI opened an entirely new window for her. She can now listen to thousands of books independently and uses AI tools to enhance images for her photography work. The technology altered not just her professional capabilities but her sense of autonomy as a student and working professional.

A similar journey unfolded for Zahra Ali Al-Safar, a handicrafts instructor at the Madhar Charitable Society for the Care of Blind Women in Qatif.

After learning traditional crafts and becoming a recognised trainer at several charitable organisations, she discovered AI applications that help her assist blind trainees in creating marketing campaigns for their products.

The technology also allows her children to complete their homework assignments.

"What bothers me most about AI is the repetition – it always repeats questions about what I'm asking for," Zahra notes, highlighting the technical limitations that still frustrate users despite the revolutionary capabilities.

These individual stories represent a broader demographic change. In Egypt, the 2022 National Survey of Persons with Disabilities found that the prevalence of visual impairment ranges from 1.41% (approximately 1.5 million people) to 8.38% (more than 9 million people), depending on whether the "moderate disability" or "broad disability" definitions are used.

By January 2023, Ahmed Al-Maragi, the editor-in-chief of Al-Akhbar Braille magazine, estimated that Egypt's blind population was 1.8 million.

Professional breakthroughs and persistent barriers

Ahmed Hossam Al-Karani, founder of the Blind Compass team and trainer in Egypt's presidential Mawadda project, exemplifies how visually impaired professionals are pushing AI applications beyond conventional uses.

Ahmed employs AI not only for typical tasks like research, translation, and summarisation, but also to monitor Egyptian stock market indicators and analyse market movements.

Traditional screen readers cannot directly interpret numerical indicators, but AI enables them to recognise numbers and identify optimal moments to buy and sell stocks. This application demonstrates how assistive technology can open entirely new professional pathways for visually impaired individuals.

"Such tools can help bridge a clear economic gap, since employment opportunities available to visually impaired people don't exceed five percent, and workers often receive wages below the legal minimum, worsening their living conditions," Ahmed explains.

Dr Ahmed Rahma, an assistant professor at Cairo University's Faculty of Media and the first visually impaired teaching assistant in his department, uses AI tools for academic research but encounters significant obstacles.

Audio transcription tools fail to accurately recognise local dialects and produce incoherent texts even when working with standard Arabic content.

"Another problem is the appearance of inaccurate research results, where these tools assign titles to books or names to authors that don't exist in reality, which confuses me as a researcher and forces me to turn to alternative sources," Dr Ahmed explains.

The cost barrier presents another significant challenge. Most specialised academic research applications impose monthly subscriptions ranging from $20 to $200, amounts that Egyptian researchers find difficult to afford, especially when needing multiple tools simultaneously.

Mohammed Ibrahim, a psychological and family counselling specialist, faces restrictions when researching sensitive topics through AI platforms.

The tools impose strict limitations on searches involving subjects like sexual orientation, refusing to engage with specific terminology or forcing him to modify his queries until the system responds.

"This obstacle consumes a lot of time in repeated attempts to get an answer, and can take hours of continuous research," Mohammed says, describing how content filtering undermines legitimate professional research.

The accessibility challenges extend beyond individual frustrations to reveal systemic inequalities in how AI development prioritises different markets and languages.

These tools cannot browse Egypt's Knowledge Bank or well-known global digital libraries, depriving users of rich scientific resources that could save significant effort.

Mohammed Tawfiq Bellow, a writer and former Secretary-General of the Ibsar Association in Saudi Arabia, emphasises that AI represents "the coming scientific revolution," comparable to the impact of computers when they first appeared.

However, he stresses the need for intensive training programmes to enable people with disabilities to benefit from this technology across all aspects of their daily lives.

Mohammed warns against over-reliance on AI as a substitute for developing personal skills or compensatory senses, which could create risks that limit individuals' real capabilities. He advocates clear goal-setting when using AI and assistive technologies, followed by the acquisition of necessary skills for effective use.

Balancing promise with caution

Religious content creator Sheikh Abdulrahman Ibrahim raises concerns about AI's impact beyond academic research, particularly regarding religious and historical beliefs. He fears that religious texts or scripts created through AI might be recorded and published without accuracy verification, potentially contributing to the spread of misinformation.

"What really scares me is that religious texts or scripts might be written via AI, then recorded and published without verifying their accuracy, which could contribute to circulating incorrect information," Abdulrahman explains.

Despite acknowledging AI's value in content creation through text summarisation and image design and editing, Sheikh Abdulrahman emphasises the importance of keeping it as a supporting tool rather than a final source of knowledge.

The technology's limitations become particularly apparent when dealing with Arabic language content and regional databases.

Most AI tools struggle with Arabic dialects, cannot access local knowledge repositories, and often reflect biases embedded by their primarily Western developers.

Amjad advocates for more advanced technologies to become available in Egypt, such as AI-powered glasses, at prices accessible to all blind individuals. Such developments could help address the significant employment gap facing visually impaired communities, where job opportunities remain severely limited and wages often fall below legal minimums.

Users like Khalid Al-Harbi, Amjad Al-Mutairi, and Ahmed Al-Karani have gained greater independence through AI applications for reading, navigation, and professional tasks.

However, they also face limitations, including language support gaps, cost barriers, and content filtering restrictions, which affect their work.

The experiences documented reveal challenges in AI development for Arabic-speaking communities. Technical issues include poor recognition of local dialects, limited access to regional databases, and subscription costs that may be prohibitive for users in the region. Some users report having to switch between multiple applications to access free alternatives when paid services become unaffordable.

These user experiences highlight ongoing questions about accessibility, language diversity, and economic factors in technology development.

The extent to which AI tools can address professional and educational needs for visually impaired individuals will likely depend on how developers and policymakers respond to current technical and economic barriers.

Shimaa Elyoussef is a freelance Egyptian journalist

This piece is published in collaboration with Egab