Appendices

Appendix 1: Overview of Ethical AI practices by Eman Ahmad Al-Kuwari, Director of Digital Innovation Department, Ministry of Communications and Information Technology

1. Ethical AI Principles and Frameworks that guide AI practices in line with our culture:

National regulators must implement nationwide frameworks and ensure their adoption. These frameworks should consider ethical principles and guidelines that align with Arabic values such as transparency, fairness, accountability, and respect for individual privacy. Private organizations should also ensure that the right policies are set in place to ensure the safe practice of AI that takes into account Arab culture.

Example: Qatar's National AI Strategy is focused on six pillars: education, data access, employment, business, research, and ethics. This represents a key step toward ethical AI practices at the national level. Further initiatives and artefacts endorsing safe AI practices are expected.

2. Inclusive AI solution development

Companies developing AI solutions should avoid biases that may disproportionately impact certain groups within the Arabic-speaking population. Thorough studies should be conducted to identify and include all target groups within the solution. Ensure that AI technologies are developed with diverse teams that include individuals from different cultural backgrounds. This approach helps in creating more inclusive and culturally sensitive AI systems.

It is also crucial to prioritize the protection of individual privacy when developing AI-powered solutions, aligning with our cultural values. Hence, it is important that AI solutions are developed in line with the National Personal Data Protection Law and international best practices.

Additionally, encouraging localized solutions where local startups tailor AI applications to address specific needs and challenges within Arabic-speaking communities is essential. Consider the unique linguistic, cultural, and social aspects when developing AI systems.

3. Increase education and awareness:

Promote AI literacy and educate the public about AI to foster a better understanding of its capabilities and limitations. Address misconceptions and promote awareness of the potential benefits of AI aligned with Arabic values.

Example: Several initiatives have been launched by MCIT in relation to Digital Skills. These initiatives aim to underline the impact of emerging digital technologies and trends on the current and future workforce of Qatar. With the support of key ecosystem stakeholders, the Ministry is working towards integrating the latest technological developments into education and training curricula. For more information, visit the Artificial Intelligence Committee page on the Ministry of Communication and Information Technology website.

4. Build Trust and Communication:

As a nation, foster trust by maintaining transparent communication about AI development and its impact. Address concerns and provide clear information about the ethical considerations and safeguards in place. Open dialogue and communication channels are crucial for building and maintaining public trust in AI initiatives.

Appendix 2: Artificial intelligence in the life of people with disabilities by Dr Hayat Khalil Heji

1. How AI is being used by the blind community today:

1. Image recognition techniques to convert visual information into read recognizing
2. Automated reading systems to convert written text into audible speech.
3. Personal assistance applications that use artificial intelligence technologies to assist with everyday tasks such as navigation and object recognition.
4. Location identification techniques using artificial intelligence to guide people with visual impairment during navigation and location identification.
5. Technologies for detecting obstacles in the path and directing people with visual impairment to avoid them, which enhances safety during transportation.

2. The impact of artificial intelligence on the life of people with disabilities:

People with disabilities use artificial intelligence devices and applications, which enhance their independence and enable them to participate effectively in society. Through these modern technologies, it is possible to improve their access to health, educational and technological services, taking into account preserving the privacy of people with disabilities. Examples of employing people with visual impairment include artificial intelligence techniques:

1. Image recognition techniques to convert visual information into readable text.
2. Automated reading systems to convert written text into audible speech.
3. Personal assistance applications that use artificial intelligence technologies to assist with everyday tasks such as navigation and object recognition.
4. Location identification techniques using artificial intelligence to guide people with visual impairment during navigation and location identification.
5. Technologies for detecting obstacles in the path and directing people with visual impairment to avoid them, which enhances safety during transportation.

3. Special education can benefit from artificial intelligence in the classroom:

Artificial intelligence helps provide personalised educational experiences for each student with disabilities to suit their individual learning styles, interests, abilities, and needs. One of the applications of artificial intelligence in education is the virtual teacher supported by artificial intelligence. Employing artificial intelligence provides individual educational support to every student with disabilities that suit his abilities and needs by providing unique solutions, customising educational programs, accurately evaluating students’ progress, and identifying areas that need additional attention and development. The process of distance learning and providing students with daily life skills. Some technical challenges must be overcome, the most important of which are:

1. The integration of artificial intelligence technologies into the private learning environment and the extent of its compatibility and stability.
2. The rights to privacy, security, and confidentiality of information when collecting and analysing personal data of students with disabilities.
3. Providing intensive training for teachers in modern technological techniques.
4. Providing access to technology and ensuring its equitable availability to students in all educational settings.
5. Motivating students and encouraging them to participate in the learning process through technology.

4.The contribution of research and academia to the accessibility of artificial intelligence technologies to people with disabilities:

Research can contribute to improve accessibility for people with disabilities and enable them to use modern technologies and analyse data, taking into account the variation in capabilities and needs among people with disabilities through:

1. Incorporate the principles of comprehensive design integration into developing these intelligent technologies.
2. Enhancing machine learning better to understand the needs and preferences of people with disabilities.
3. Develop and design assistive tools that use artificial intelligence technologies to support people with disabilities in their daily lives.
4. Develop human-computer interaction systems that effectively respond to the movements and notifications of people with disabilities.

5. Strategy for the future of artificial intelligence:

1. Develop a future strategy for artificial intelligence in line with religious, cultural and social values and the Arabic language.
2. Encouraging cooperative research between universities and communities of people with disabilities, including organisations, institutions, and centers that provide services to people with disabilities, to identify and better understand their challenges and ensure that their actual needs are met.
3. Raise awareness about the importance of access to these modern technologies and train community members, especially developers and designers, on how to incorporate universal design principles.

To contact Dr Hayat please email her at: hkheji@alnoor.org.qa

Appendix 3: How AI is being used by Step-by-Step Center for Persons with Disabilities by Nasreen Sharif, and Olcay B.Connor

AI offers autistic students non-threatening, non-judgmental support. Therefore, our students are very comfortable interacting with technology.

Gaming offers a predictable, consistent environment: if they are unsuccessful in a gaming situation, they start the game again. Also, robots for instance lack humanness, which is also a positive trait especially for a person with Autism, as there are no confusing gestures or facial expressions just a simple wave or smile. This means that students can easily read those cues and focus on their learning, rather than trying to decode behaviors.

AI-powered robotic devices help our students in performing exercises and tasks that may be difficult or impossible for them to do on their own. For example, robots help our students copy simple body movements and can be used as a role model. SEN teachers use AI such as Magic School to create personalised learning plans (Individualized Education Programs or IEPs) to enhance their teaching methods and improve student outcomes taking into account different learning styles and paces. They use learning apps and software (with Smartboards or iPads) through interactive and game-like interfaces that range from reading comprehension, math’s and sciences.

Speech and language intervention: Our therapists can gather lots of data from various sources, such as patient records, therapy session recordings, and progress reports, to create tailored treatment plans that address the specific needs and goals of each individual with the help of AI. This personalized approach can lead to more effective therapy sessions and better overall outcomes.

Occupational Therapy Intervention: We utilise chatbots to develop Sensory integration programs treatment goals with just a few clicks. They also use AI-powered interactive smart board games to improve motor coordination and development. Therapists utilise chatbots to develop Sensory integration programs and treatment goals with just a few clicks. They also use AI-powered interactive smart board games to improve motor coordination and development.

Life skills: AI helps develop essential fostering activities of daily and independent living (E.g. self-care and hygiene) , and vocational skills (E.g. interview skills). For instance, our students regularly visit the AI Smart market to practice shopping & money management skills.

We use AI to convert educational materials into more accessible formats, such as large print and contrast-colored backgrounds, making content more inclusive for students with visual or sensory learning differences.

Hirfa Production Unit: Within our innovative Arts and Crafts studio, we enable our learners to create and sell their own products and crafts. AI tools like Canva & Magic Studio help our students to visualize merchandise and develop digital artwork.

Appendix 4: Interview with Mark Hughes, Executive Director of QF Special Schools and SEN Services and Dr. Imad Deeb, Assistant Director, The Learning Center (TLC)

Question:

How is AI impacting the lives of persons with disabilities?

Answer:

Mr. Mark - Overcoming learning disabilities has long been a challenge for educators and students alike. Traditional methods of teaching often fail to cater to the unique needs of students with learning disabilities, leaving them struggling to keep up with their peers. However, recent advancements in artificial intelligence (AI) technology have shown promise in revolutionizing the way we approach education for students with learning disabilities. By harnessing the power of AI, educators can create personalized learning experiences that cater to the specific needs of each student, ensuring that no one is left behind.

Dr. Imad (TLC) - By providing assistive technologies, such as speech recognition, computer vision, and natural language processing, to enhance accessibility. These technologies empower individuals with disabilities to navigate the digital world, communicate effectively, and participate more fully in society.

Question:

How is Special Education taking advantage of recent technological developments?

Answer:

Mr. Mark - One of the most significant ways AI can assist in overcoming learning disabilities is through the use of adaptive learning systems. These systems use algorithms to analyze a student’s performance and adjust the learning content and pace accordingly. This personalized approach allows students with learning disabilities to progress at their own pace, without feeling overwhelmed or left behind. Furthermore, adaptive learning systems can identify areas where a student may be struggling and provide targeted support to help them overcome their challenges.

Another way AI can support students with learning disabilities is through natural language processing (NLP) technology. NLP enables computers to understand and interpret human language, allowing AI-powered tools to provide real-time feedback on a student’s written work. This can be particularly beneficial for students with dyslexia, who often struggle with spelling, grammar, and sentence structure. By providing immediate feedback, students can identify and correct their mistakes, leading to improved writing skills and increased confidence.

Question:

In the classroom setting and what are the perceived challenges or opportunities?

Answer:

Mr. Mark - One of the most significant limitations of artificial intelligence is that it can be wrong or unreliable in its predictions or solutions. AI algorithms are only sometimes easy to follow, making it difficult for marketers to understand how an AI prediction or solution came to be. Not understanding or contributing to this process makes it harder to trust the information returned to you. This means AI designers must fact-check and ensure the predictions and solutions provided are accurate and reliable. Not only does this bring up issues of transparency, but it also raises concerns about fairness and accountability. Schools using AI must ensure that the AI algorithms they rely on are transparent and explainable to all stakeholders. By doing so, you can build trust and confidence within your team when using these AI-driven tools, boosting your overall productivity and streamlining your workflow.

It’s also important to remember that AI systems need instructions, adjustments, and controls need to be inputted by humans to run successfully. AI tools can generate much content and produce predictions, but this all relies on specific human-generated prompts. Despite being clear and understandable to humans, these prompts may also generate incorrect information. This can be due to AI’s inherent limitations, such as quick content proliferation and narrowed access to resources (i.e., databases, updated internet archives, etc.).

Dr. Imad (TLC) - Adaptive learning platforms, virtual reality, and AI-driven personalised learning tools. These technologies cater to diverse learning needs, offering personalized instruction and real-time feedback, but these need teachers training and there are concerns regarding privacy.

Appendix 5: Statement from Dr. Mohamed Koutheair Khribi, Head of Training Section at MADA

MADA Academy is an initiative recently launched by MADA Center toward empowering people and institutions, through providing engaging and inspiring inclusive training and capacity development for all, as well as supporting open and inclusive digital education IDE services.

MADA Academy serves as a center of excellence in training, professional development, and lifelong learning for all. The academy offers a large selection of technology-enhanced training activities, delivered within various training modalities. There are basically two main training tracks so far: the ICT accessibility and Assistive technology tracks, each of which covers several topics, including the effective integration of technology in education toward inclusive and accessible learning experiences for all. Visit MADA Academy through the dedicated website.

All training courses that we’re delivering are aligned to specific competency frameworks MADA ICT-AID competency framework. MADA ICT-AID competency framework is available through their dedicated MADA ICT Aid website. This open competency framework describes all the relevant ICT accessibility competencies and abilities required for everyone to be able to use and develop accessible content, services and products. MADA framework can be used to assist professional education services, universities and individuals on delimiting the required relevant competencies in the ICT accessibility field and fostering the integration of ICT accessibility courses in universities curricula and training programmes. As an open framework, it can be adapted for use in different learning contexts and modes, and availed to develop, describe and publish ICT-AID aligned resources in courseware repositories. It is worth noting that MADA ICT-AID is also featured as an educational standard available to users of the well-known OER Commons digital library and collaboration platform. It can be used therefore to index and align ICT-AID educational resources providing accordingly ease of access and retrieval of Open Educational Resources OER serving global learners and educators.

Furthermore, in order to address the lack of availability of ICT-AID aligned courses worldwide, we have recently launched a dedicated open access repository: “MADA ICT-AID OER Hub” intended to be e a Global knowledge hub featuring freely accessible resources on ICT accessibility. The Hub is gathering collections of open educational resources, aligned to MADA Framework. All resources are aggregated, curated and managed by MADA and partners, through collections, and groups, and development tools available on the Hub. The community of ICT accessibility professionals, experts, advocates, educators, and learners can discover, create, and share open learning and teaching resources, and connect with others to expand their capabilities and improve inclusive practices.

We have been collaborating with partners, namely, the Ministry of education and higher education (TEDC) and universities in Qatar (CCQ, HBKU, UDST) to develop digital accessibility courses aligned with MADA ICT-AID competency framework and to integrate these courses in undergraduate (CCP) and postgraduate levels.

For the time being I am trying to integrate AI capabilities into ICT-AID assessment.

To learn more about MADA ICT-AID Competency framework, visit the MADA ICT Aid website.

To learn more about MADA OER HUB, visit the open-access repository available through the MADA ICT-AID OER Hub website.

Here is a list of four of our related publications:

• M. K. Khribi. (2022). MADA ICT Accessibility and Inclusive Design ICT-AID Competency Framework. Nafath, 7(21). https://doi.org/10.54455/mcn.21.04
• M. K. Khribi, A. Othman, A. Al-Sinani. (2022). “Toward Closing the Training and Knowledge Gap in ICT Accessibility and Inclusive Design Harnessing Open Educational Resources”. The 22nd IEEE International Conference on Advanced Learning Technologies ICALT 2022.
• M. K. Khribi, A. Othman, A.N. Al Jabor. (2022). “Fostering ICT accessibility proficiency through Mada ICTAID Competency Framework”, The 8th International Conference on ICT & Accessibility (ICTA), 2021.
• M. K. Khribi, & A. Al-Sinani. (2021). Harnessing OER to build capacity in ICT Accessibility and Inclusive Design. Open Education Global Conference, OEGlobal’21.

Appendix 6: Statement from Disability Activist, Ms. Ghanimeh El-Taweel

1. So, let's start with the information pack. The information pack had all the information about this session. However, it is not that accessible. An example, the colors used are not high contrast. It is blue on blue, and the text was not clear. I will leave it to the accessible document experts to educate us more on what else is needed.
2. I think the fact that this discussion was held in English and not Arabic is something to think about. Culture was brought up in the roundtable. Culture and language go hand in hand. The official language of the country is Arabic, and we need and want to encourage more Arabic responses and participation.
3. We actually are at a point where we cannot discuss disability or accessibility in just Arabic because Arabic discussions are so limited that we do not have proper terminology to address these issues.
4. I think the topic should have been more specific to Qatar. What is available in Qatar and how can Qatar improve? What is needed in Qatar? What is working and what is not working?
5. The whole idea of a roundtable discussion is good and well. But as Dr Josélia said in her introduction, it is now time to walk or even run. However, my experience with roundtables in Qatar so far have not gone past the discussion. I think it would have been better to form a task force and decide on what are the actions that needed to be taken and start that discussion. The information pack said there would be 10 key actions, but correct me if I am wrong, not even one was decided.
6. The terminology used in the discussion was quite problematic. Academics were using terms like normal vs special needs. This is completely offensive, especially when they referred to the end users who have disabilities as a niche and boiled everything down to does this make us money or not. One of the key words in the discussion was humanity. Is this humanity? Plus, as Dr Camilla said, it is an untapped market and as another said it is not two different apps it is one app that benefits all.
7. It was mentioned that AI is not judgmental. However, those of us who are Arab or Muslims are aware of how biased AI is. First, AI is not very compatible with the Arabic language, nor does it recognize specific linguistic and dialectal features. Plus ask AI about Palestine and it will show you how biased it is.
8. The topics discussed went beyond AI, which is extremely important. AI is just a tool and at the moment the foundation is lacking
9. A thought leader brought up legislation. I could be wrong but how is it legislation when it is specific to the governmental entities and no one else? Doesn't legislation mean there is a law that all must follow? Why is the legislation specific only to websites? Is that the only online platform that this covers?

To contact Ghanimeh please email her at ghanimeh@gmail.com

Appendix 7: Statement from Leyla Mroueh, Disability Advocate and Public Affairs Manager at Sasol

I am a person with an 'invisible' disability who has been employed in creative media environments, such as advertising, TV, and radio, for over 20 years. I never considered myself as having a disability as I have been fortunate to be surrounded by people who have made reasonable adjustments, allowing me to participate and contribute fully. In my work in broadcast media, and subsequent work in both the public and private sector, I have had the freedom to challenge traditional ways of working and processes to build more horizonal dot joining to drive human -centric warm webs of knowledge sharing, and create inclusive and equitable ecosystems, programs and initiatives that support organizational objectives and drive social impact.

This was and still is possible because of people, not technology. For my own experience to be the norm It was also a corporate culture of equity and inclusion, driven by organisations like Qatar Foundation and Sasol that champion diversity and have the institutional courage to think outside the box. For other, it may be the families who are informed and schools that embrace customized learning and other less tangible enablers that are the ‘difference’ that I believe ‘make the difference’ in any policy solutions ahead. With the increasing dependency on technology, AI has added value to my life by providing additional writing support through programs like Grammarly, and it has allowed me to personalise my educational journey through various platforms. However, AI is not perfect, and I still face challenges with software that is not user-friendly for neurodiverse individuals or digital services that have been made for more neurotypical users. I look forward to further advances as we build more multi-sectorial bridges and join new dots between people, resources and impact and encourage more dialog and debate.

Recommendations:

1. We must view AI's potential holistically, in context with other human-centric enablers, equitable and inclusive practices, and reasonable adjustments to support individuals with disabilities in the future of work if we want to stand a chance of building a truly equitable and inclusive future for all.
2. We must ensure that technical equity and affordability are considered for any product or services. This can be achieved by having reduced rates or lending libraries from which people can loan equipment.
3. Continue to spotlight senior leadership publicly supporting the importance of Ethical AI and championing it, celebrating startups and organisations working towards contributing to it in line with the 2030 vision.
4. We need stronger collaboration and coordination at a national level to ensure that laws and policies ensuring inclusive hiring practices, events, and activities are followed. There should be consequences for non-compliance.
5. It is important to raise awareness that not all disabilities are visible. This can be achieved through positive media spotlighting and role models that challenge perceptions and empower persons with disabilities.
6. We must work with families and schools to encourage more empowered thinking. This means acknowledging that everyone has a place to contribute to the future of work. In many cases, it’s just a matter of ensuring that the right enablers are in place, from inclusive hiring practices to matching skills to opportunities.
7. Leaders and HR teams need more training to ensure that they can see the benefits to businesses of having a fully inclusive and diverse workforce.
8. Greater media representation of persons with disabilities is important. Working closely with key stakeholders, this is very possible and will inspire others, as if you can ‘see it’, you can ‘be it’!

To contact Leyla, please send an email to: leyla.mroueh@sasol.com